• Home
• Preliminary Program

10:00   Poster session 2 (Odd numbers) 10:00 0 mins Detection of freezing of gait and fear of falling in daily life Juan Delgado Terán, Laurens Kirkels, Tjitske Heida, Richard van Wezel Abstract: Freezing of gait (FoG) has been defined as a brief episode where the subject cannot move forward despite the intention to walk. FoG often leads to balance impairments and constitutes a frequent cause of falling in patients with Parkinson’s Disease (PD). FoG frequently coexistence with non-motor symptoms such as depression, anxiety, and Fear of Falling (FoF), which are the strongest predictors of low Quality of Life. Assessment of most PD symptoms is currently performed via questionnaires, and this might lead to biases. However, most people with PD are unable to correctly identify FoG, leading to underdiagnosis. Accurate and continuous evaluation of FoG in home environments using wearable sensor data is critical to diagnose and adequately managing FoG. However, most systems have failed to deliver the same performance in home environments because algorithms created in the laboratory do not address the large variability of human behavior seen in daily life (free‐living conditions). Our main goal is the detection of FoG and FoF using Machine Learning algorithms including physiological sensors (PPG (e.g., heart rate variability), EEG, and skin conductivity) and wearable movement sensors (pressure-measuring insoles and inertial measurement units (IMUs) under semi‐controlled, semi‐free living, and free‐living conditions. The configuration of the sensors placed on the person with PD is essential to predict FoG and FoF. Moreover, the obtained dataset will be used to train novel algorithms to investigate the influence of different types, numbers, and locations of sensors on the performance of FoG and FoF detection. Moreover, with the optimized measurement configuration and machine learning algorithms, we aim to detect and even predict FoG episodes in semi-controlled and free-living conditions. 10:00 0 mins Assessing the impact of arm weight support and tactile stimuli on weight perception in robotic upper limb rehabilitation Alexandre Ratschat, Gerard Ribbers, Laura Marchal-Crespo Abstract: Neuroscientific evidence suggests that people after stroke should engage in task-specific, high-intensity training to maximize their recovery [1]. Additionally, somatosensory information plays a fundamental role in generating skillful movements [2]. Current rehabilitation robots do provide high-intensity training, however only for non-functional movements and with limited simulation of somatosensory information [3]. As part of the HyperRehab project, we aim to simulate realistic tangible virtual objects to effectively train activities of daily living using an upper-limb exoskeleton and immersive virtual reality. Here, we focus on the users’ perception of simulated objects in combination with robotic assistive strategies like arm weight support. We plan to assess the weight perception for four conditions: (1) Lifting real weight, (2) lifting weight simulated by an upper-limb exoskeleton, (3) lifting real weight with arm weight support from an exoskeleton, and (4) lifting weight simulated and arm weight supported by an exoskeleton. This study will be performed with 24 healthy participants. Using a two-alternative forced choice and an up-down transformed response method we will determine the just notable difference (JND) of weight perception for the four conditions while monitoring the muscular activation using Electromyography. We expect the JND for the real weight conditions to be lower than the JND of the simulated weight conditions since the robot currently has no capabilities to render tactile stimuli related to weight perception. For the assistive arm weight support conditions the assistance and weight simulation signals are rendered by the same actuators of the robot, potentially leading to hampered weight discriminability, i.e. higher JNDs. Finally, we will analyze how muscle activation differs per condition and whether this correlates with the perception of weight differences. The results of this study will help us to improve our understanding of tangible virtual object perception in upper-limb rehabilitation. Based on our findings, we will propose guidelines to enable realistic perception of tangible virtual objects, leading to improved transfer from training in the clinics to daily life and ultimately improving the quality of life of people after stroke. 10:00 0 mins Towards computer aided design of passive heat and moisture exchangers Maartje Leemans, Maarten van Alphen, Wim Vallenduuk, Richard Dirven, Michiel van den Brekel, Saar Muller Abstract: Background: Small passive Heat and Moisture Exchangers (HMEs) are standard treatment for pulmonary rehabilitation after a total laryngectomy. These HMEs consists of a plastic cassette containing a foam material coated with hygroscopic salt (wet core), which acts as a condensation and evaporation surface. During each breathing cycle, part of the moisture from the exhaled air is stored on the wet core and used to humidify the inhaled air (water exchange). Currently, the passive HMEs’ performances are not as high as the upper airways before laryngectomy, and the influence of the HME design on its performance is difficult to predict. We developed a numerical HME model to improve the physical understanding and design of these small passive HMEs. Methods: The numerical HME model was implemented in Matlab 2020b (Mathworks Inc. USA). The physical processes inside the HME are described by four discrete physical equations for the conservation of mass and energy in the air and wet core. The HME model was tuned and verified with experimental data sets of the HME’s performance (water exchange and temperature data) at standardized tidal volumes and flows. Finally, the HME model was applied to HME design variations and environmental conditions outside the scope of the experimental data. Results: Verification of the model’s results to the experimental data shows that the tuned model yields reliable results. The physical understanding as found using the model leads to the following design recommendations: the mass wet core, determining the HME’s total heat capacity, is the driving factor of the HME’s performance. Increasing the HME’s diameter is an effective way to improve an HME as it yields a higher performance and lowers breathing resistance. Furthermore, the model simulations led us to the insight that hygroscopic HMEs are not passive but are actually “reactive” to its environmental conditions. Therefore, HMEs intended for use in warm or dry climates should contain more and those for use in cold humid climates less hygroscopic salt. Conclusions: The numerical HME model improves the physical understanding of small passive HME’s, predicts the HME’s performance and aids the developmental process towards better performing HMEs. 10:00 0 mins Unobtrusive wearable sensing to estimate human circadian process Nemanja Cabrilo, Charikleia Papatsimpa, Jean-Paul Linnartz Abstract: The central biological clock in the brain has a near-24h rhythmicity that is a main determinant of individuals’ sleep/wake cycles. It also orchestrates the daily rhythms of hormonal secretion and behaviour such as subjective alertness and performance[1]. Having a solution to determine and track the state of an individual’s internal circadian state can allow us to make major steps to optimize human daily rhythms, from the use of light for improving sleep quality to precision medicine, diagnosing neurological disorders and optimizing drug delivery. Some estimates[2] of the circadian state demonstrated that, even if not up to clinical standards, can greatly enhance human-centric lighting to improve wellbeing of, for example, office, industry or warehouse workers. In a quest to investigate the use of wearable sensing for tracking the underlying circadian process, we developed a model that combines a physiology-based model of the human biological clock with non-invasive but possibly in-accurate ambulatory data (in particular actigraphy data) in a statistical framework[3]. Nowadays we are conducting a field study to validate our model in a real-life setting and to test to what extent this leads to meaningful estimates. We compare model-based predictions based on wearable data (CamnTech MotionWatch8) against the “gold standard” circadian state estimation, namely, individual subject’s bathyphase, or timing of the daily Core Body Temperature (CBT) nadir (minimum point) which is a widely accepted circadian phase marker[4]. CBT is continuously being recorded with BodyCAP e-Celsius Performance ingestible e-capsule(s). Beyond, we aim to assess further physiological signals and identify the best possible predictors, with reasonable user comfortability, as non-invasive circadian biomarkers, for instance: Skin temperature (DSL1922L temperature logger), Heart-Rate Variability (Movisense EcgMove4), Electro-Dermal Activity (Empatica E4 wristband). Accurate and unobtrusive estimation of the exact circadian phase can unlock the potential of numerous applications, possibly including personalized Human-Centric Lighting. Our measurement set-up is unique in its versatility of different sensor modalities and allows a cross-benchmarking. As measurements are being collected at the moment of writing this abstract, the conference paper will primarily address design consideration for the tests, discuss variability of measurements and sensor imperfections combined with early experiences and technical findings. 10:00 0 mins Sleep characteristics in patients with COMISA compared to OSA and insomnia Bernice Wulterkens, Lieke Hermans, Pedro Fonseca, Jerryll Asin, Nanny Duis, Sebastiaan Overeem, Merel van Gilst Abstract: Study objectives: Obstructive sleep apnea (OSA) and insomnia frequently co-exist, which complicates treatment, for example resulting in a worse compliance to continuous positive airway pressure (CPAP) therapy. Early recognition of comorbid insomnia and sleep apnea (COMISA) is important since both sleep disorders can aggravate each other and COMISA is associated with increased risk of all-cause mortality. Research is needed to identify characteristics that can be used to recognize patients with COMISA, in order to optimize treatment approach from the start. The aim of this study is to investigate differences in sleep structure between patients with pure OSA, pure insomnia and COMISA. Methods: We obtained polysomnography data from 326 patients from the SOMNIA database. The group included patients with OSA (n=199), insomnia (n=47) and COMISA (n=80). We compared statistics related to sleep stages, awakenings and sleep disordered breathing (SDB) events between the three patient groups. Results: Wake after sleep onset (WASO) was significantly longer for the COMISA group compared to OSA (median, 83.3 vs 60.0 minutes, p<0.01). No significant differences were found in the total number of awakenings and the number of short (up to and including 2 minutes) and medium-length awakenings (2.5 up to and including 4.5 minutes). The number of long awakenings (five minutes or longer) and WASO containing only long awakenings were significantly higher for the COMISA group compared to OSA (median, 3.0 vs 2.0 awakenings, p<0.01, and median, 43.3 vs 25.5 minutes, p<0.001). The apnea-hypnopnea index (AHI) and the 4% oxygen desaturation index in COMISA were -as expected- higher compared to insomnia (median, 16.6 vs 6.8 events/hour, p<0.001, and median, 4.5 vs 0.8 desaturations/hour, p<0.001), but lower compared to OSA (median, 16.6 vs 21.6 events/hour, p=0.0136, and median, 4.5 vs 8.5 desaturations/hour, p=0.01). Conclusions: Patients with COMISA seem to present with a milder form of OSA, but nevertheless presented a more disturbed sleep structure, mainly characterized by prolonged awakenings. Further research is needed into these sleep characteristics that may lead to a better understanding of the mechanisms involved, earlier diagnosis and better treatment strategies. 10:00 0 mins An In Silico clinical trial on Coronary fractional flow reserve as a replacement for the original clinical trial: a feasibility study Pjotr Hilhorst, Wouter Huberts, Rajarajeswari Ganesan, Pim Tonino, Frans van de Vosse Abstract: In Silico clinical trials have great potential for replacing clinical trials. In this study, we aim to demonstrate the feasibility of conducting in Silico clinical trials by generating virtual patients and reproducing a clinical trial in which the clinical benefit of fractional flow reserve (FFR) measurements was demonstrated (i.e., the FAME study [1]) for patients suffering from coronary artery disease. Here, we will present the strategy we envision to demonstrate the clinical benefit of the FFR using in Silico trials only. In addition, we will present preliminary results regarding model development. A one-dimensional pulse wave propagation model (PWPM) that is capable of computing patient-specific FFRs has been developed. Sensitivity analysis will be conducted to prioritize parameters for model personalization. Geometric information will be extracted from angiograms, whereas patient-specific parameters will be estimated by using a machine-learning model that is trained using angiograms, demographics, and pressure losses across stenoses. The latter will be based on FFR measurements or 3D computational fluid dynamic simulations. A synthetically generated training set will be used to assure a large enough dataset and sufficient coverage of the heterogeneity within the population. Secondly, the parameters can be varied to generate virtual patients. In the future, the model output will be transformed into a clinically relevant output (i.e., mortality and morbidity) through a transfer function. Furthermore, the approach will be evaluated on an independent set of real clinical trial data. The results showed that the PWPM can accurately model coronary pathophysiology. The computed FFR values were plausible compared to clinical findings, which typically show an FFR below 0.8 around a stenosis severity of ≥ 50/60% [2]. Overall, it could be concluded that this model is ready for the next step in our devised methodology and can be used as a virtual cohort generator to recreate the FAME1 study in Silico. 10:00 0 mins NURBS-based isogeometric analysis of a bi-ventricular heart model Robin Willems, Clemens Verhoosel, Olaf van der Sluis Abstract: Patients suffering from Ventricular Tachycardias (VTs, a fast abnormal heart rate), require ablation therapy to treat the scar tissue located on the left ventricle which causes VTs. Success rates are considered moderate (50-80%) and can be elevated using patient-specific three-dimensional computer models of the human heart, specifically the (bi-)ventricles. Current electromechanical ventricular models using Finite Element Analysis (FEA) are computationally expensive, which is impractical for clinical applications. Furthermore, patient-specific FEA models rely on accurate input data to produce meaningful results. We propose an approach in which the Isogeometric Analysis (IGA) paradigm [1] is applied to existing mechanical ventricular models [2] subject to limited available input data. IGA enables the construction of smooth discretized geometries, which require fewer degrees of freedom, making them ideal for clinical applications. In this project, we present a method for constructing a template bi-ventricle NURBS geometry comprised of two truncated ellipsoids, using multipatches. We target the mapping of the template geometry onto patient-specific ultrasound scan data. The cardiac mechanics and zero-dimensional circulatory system are solved within the IGA framework through a monolithic strategy. The results are compared to an existing model which employs the traditional finite element approach. The novelty of our research resides in the implementation of IGA for a patient-specific cardiac model that is based on scan data. IGA is expected to reduce the computational effort [3] when compared to the currently used methods in the literature while enabling an accurate representation of the cardiac mechanics. [1] Hughes, T.J., Cottrell, J.A. and Bazilevs, Y., 2005. Isogeometric analysis: CAD, finite elements, NURBS, exact geometry and mesh refinement. Computer Methods in Applied Mechanics and Engineering, 194, 4135–4195. [2] Pluijmert, M., Delhaas, T., de la Parra, A.F., Kroon, W., Prinzen, F.W. and Bovendeerd, P.H.,2017. Determinants of biventricular cardiac function: a mathematical model study on geometry and myofiber orientation. Biomechanics and modeling in mechanobiology, 16, 721-729. [3] Pegolotti, L., Dede, L. and Quarteroni, A., 2019. Isogeometric analysis of the electrophysiology in the human heart: Numerical simulation of the bidomain equations on the atria. Computer Methods in Applied Mechanics and Engineering, 343, 52–72. 10:00 0 mins Computational model of the mechano-electrical interaction in ventricular tachycardias Evianne Kruithof, Matthijs Cluitmans, Peter Bovendeerd Abstract: Myocardial infarction causes cardiac tissue remodeling, which increases the risk of ventricular tachycardia (VT). To optimize therapy selection, computational models are used to study the effect of electrical abnormalities on scar-based VT risk [1]. These models neglect the effect of mechano-electrical interaction, which is explored in this project. Infarct-tissue remodels in both mechanical and electrical terms causing abnormal mechanical deformations and electrical activation patterns over the myocardium. Mechanical remodeling, like formation of stiff fibrosis, alters the strain amplitudes in and around the infarct area. Our hypothesis is that these altered strain amplitudes affect the electrical cell-to-cell communication over time [2], which results in an altered conduction velocity (CV) of the electrical pulse over the heart. This might be of importance to the development of VT, as CV decrease is known to affect VT inducibility. Therefore, in this study, a relation between fiber strain amplitude and CV is proposed. The fiber strain amplitude distribution in the chronically infarcted heart is obtained from a computational model of cardiac mechanics. This distribution is, by means of the proposed relation, transferred into a distribution of CV which serves as input for an electrical model to simulate VT. By incorporating the mechano-electrical interaction, the effect of the infarct on CV was found to extend beyond the fibrotic area, as opposed to traditional models where the effect is limited to this area only. The region of CV slowing was located in and around the fibrotic area, and VT was found to be more easily initiated. Therefore, the implemented mechanism might be promising to assess VT risk. Moreover, if long-term changes in strain are known, this might be a way to predict long-term areas of CV slowing, and thus VT risk, as well. However, further research is needed to investigate if the proposed relation correctly describes strain-induced remodeling of CV. 10:00 0 mins Unobtrusive monitoring of obstructive sleep apnea – the UMOSA project Fokke van Meulen, Raquel Pires Alves, Shuhao Que, Sveta Zinger, Sebastiaan Overeem, Sander Stuijk Abstract: Obstructive sleep apnea (OSA) is one of the most common sleep disorders. People with sleep apnea (partially) stop breathing for several seconds, multiple times per night. The gold standard for the diagnosis of OSA and other sleep disorders is commonly referred to as polysomnography (PSG). During a PSG, multiple parameters are measured to evaluate brain activity, heart rate, respiration rate, nasal airflow, body posture and limb movement. The large number of wired sensors used during a PSG may influence actual sleep quality, which may result in a less representative assessment of sleep. Several proofs of principle have demonstrated that most vital signs parameters or surrogates relevant for diagnosing OSA can be continuously measured in a fully contactless manner. Within the UMOSA project, four fully contactless measurement techniques were selected and will be further developed to be able to monitor sleep and to diagnose the presence of (obstructive) sleep apnea events. First, remote photo-plethysmography, which is the contactless counterpart of conventional pulse oximeters, using plethysmography as the measurement principle to measure instantaneous heart rate and heart rate variability. Second, speckle vibrometry uses an interferometric technique to measure cardiac and respiratory signals at the chest even when covered with one or more textile layers. Third, remote thermography measures respiration and nasal flow by aiming passive thermal radiation detectors at the nostrils and the area around the person’s head. Fourth, a high resolution near infrared camera and an infrared light source are used for the detection of body motion, body posture and respiration motion under no visible light conditions. Based on the results of laboratory research in healthy participants, the UMOSA measurement setup is installed in the center for sleep medicine Kempenhaeghe for a clinical feasibility study. Planned simultaneous recordings of the UMOSA setup and reference PSG allow the evaluation of the setup’s performance in sleep disordered patients in a clinical environment. Results should demonstrate the setup’s accuracy, coverage and feasibility to measure vital signs parameters, estimate sleep stages and its ability to detect (obstructive) sleep apnea events. Results may also point out potential redundancy within the setup. 10:00 0 mins increasing the Fixation of Spinal Bone Anchors by increasing Cortical Bone Contact Esther de Kater, Aimée Sakes, Paul Breedveld Abstract: Currently, pedicle screws are the conventional fixation method in spinal fusion surgery. However, the fixation strength of these pedicle screws may lack, preventing the desired fusion between the adjacent vertebrae and causing a need for revision surgery. Loosening of pedicle screws is a problem that is especially common in vertebrae with decreased bone density due to osteoporosis. The vertebra comprises a compact outer layer of cortical bone that surrounds the much softer cancellous bone. The conventional pedicle screw is for the major part in contact with the soft cancellous bone, only the small section of the screw located within the pedicle of the vertebra has purchase in the strong cortical bone. Although the purchase in the cortical bone is small due to the oval cross-section and the hourglass shape of the pedicle this small section accounts for 60% of the pull-out strength of the screw [1], [2]. Screw loosening is often caused by toggling, a cyclic loading that is exerted perpendicular to the screw axis causing the screw to pivot. During toggling, the cancellous bone surrounding the screw is compressed, compromising the fixation strength of the screw [1], [3]. Increasing the contact area between a spinal bone anchor and the strong cortical bone layer of the pedicle could increase the toggling resistance of the spinal bone anchor. A scaled-up two-dimensional prototype of a spinal bone anchor that can expand within the pedicle such that the anchor can shape to the hourglass shape of the pedicle was designed and manufactured. The prototype comprises ten stainless steel wedges that expand and mould to the hourglass shape of the pedicle by tensioning the central bolt. The prototype showed an increased toggling resistance compared to a non-expanding mock-up anchor. This preliminary prototype shows that an expandable in-pedicle anchor could be a feasible option to increase the contact between the anchor and the cortical bone layer, and increase toggling resistance and thus the fixation strength of spinal bone anchors. 10:00 0 mins The ability of a real time location system (RTLS) and the electronic medical record to monitor hyper-acute workflow and identify bottlenecks Igor Paulussen, Supriyo Chatterjea, Julia van Tuijl, Ben Jansen, Tom Defossez, Rick Bisschops, Gerrit Noordergraaf Abstract: Main research question: Electronic Medical Record (EMR) data are frequently used as source for quality indicators of hyper-acute clinical pathways. In stroke care, door-to-needle time (DTN, i.e. IntraVenous Thrombolysis, IVT) should be ≤60 minutes. Validity of timestamps, however, remains uncertain. RTLS technology tracks tags (patients, essential equipment and staff: nurses and technicians) within predefined areas and automatically adds timestamps. This data can be aggregated or individually assessed. Using patients presenting to the Emergency Department with suspicion of a stroke and using RTLS supplemented by EMR data, we assessed validity of times, and sought insights into process inefficiency as well as badge-wearing acceptability by professionals. Research method: A clinical, single-center prospective study using the stroke pathway with enrollment from 1-9-2020 through 31-8-2021. Ischemic stroke and IVT and/or intra-arterial thrombectomy (IAT) patients were included after informed consent. The RTLS sensors were placed in the Emergency and Radiology Departments. Patients received tags upon arrival. Interventions using RTLS input into the pathway were introduced (starting 01-07-2021) validating whether RTLS visualizes changes and that these were effective. Results: High inclusion rate was achieved (99%, n=125 patients). EMR time data proved highly inaccurate while RTLS demonstrated high fidelity. For example, maximum difference in DTN time between RTLS and EMR was 26 minutes. The overall compliance in wearing a badge was 81 ± 16%. Compliance (acceptability) was variable: nurses 96%, Neurology residents 64%, and Radiology technicians 85%. Patient tag signals not visible for RTLS (i.c. temporarily under a cover) occurred at least once in 74% of cases, typically during CT-scans. We were able to recognize and suggest adaptations for bottlenecks in the stroke pathway resulting in 25% time saving through: a) using real entry-into-workflow time; b) improving content with an IVT and IAT box, by bundling materials, thus allowing the nurse to stay in the admission rooms; and c) creating a dedicated anesthesia cart including 24/7 stand-by anesthesia care station. Conclusions: RTLS with EMR data offers a detailed, independent insight into hyper-acute care and discerns inefficient constituents. Professionals are ambivalent about badge-wearing. 10:00 0 mins Validation of a novel 3D-scanner for measuring cranial volume and head circumference of preterm infants Ronald van Gils, Linda Wauben, Timothy Singowikromo, Thomas Feijts, Crystal Wang, Bram van Deurzen, Hil Steketee, Onno Helder Abstract: Background Brain growth of preterm infants, cared for in incubators at a neonatal intensive care unit (NICU), is estimated by measuring head circumference (HC). Estimating brain growth by HC, however, has limitations. First limitation is accuracy: HC poorly correlates with brain volume. To estimate brain growth more accurately, cranial volume (CrV) measurements are preferred over HC. Second limitation concerns stress-related negative effects: HC measurements with a tape measure wrapped around the head causes stress to the infant. Aim To overcome these limitations, a novel 3D-scanner, the MONITOR3D, was developed for stress-free measurements of HC and CrV of preterm infants lying in incubators. To be stress-free, the MONITOR3D scan sessions are merged with routine care procedures. This study aims to validate the usability and accuracy of measurements of the MONITOR3D. Methods Validation took place at a NICU incubator bedspace with a prepared preterm doll lying in an incubator. Pairs of NICU nurses performed scan sessions, changing the infant’s position as they would do during routine care procedures. Also, they measured HC with the currently used tape measure, acting as golden standard. Ground truth values for HC and CrV were defined by reference scans with a commercially available clinical 3D-scanner. To estimate disturbances to the infant due to the scan sessions, sensors registered noise, light, and head-movement. In addition, nurses filled out a questionary to qualitatively assess expected disturbances. Results 56 scan sessions resulted in 25 usable scans, from which HC and CrV could be derived. The median HC of both tape measure and ground truth did not show a significant difference compared to the median HC of the MONITOR3D. The CrV measurements, however, did significantly differ from the ground truth. The scan sessions added no significant disturbances to the routine care procedures. Nurses expected the current tape measure to be more stressful than the MONITOR3D. Conclusions The MONITOR3D scan sessions can take place during routine care without causing extra disturbances. HC measurements with the MONITOR3D seem clinically accurate enough. However, CrV measurements are not accurate enough for clinical use. Before in vivo validation, further research is needed to improve CrV measurement accuracy. 10:00 0 mins Physical activity patterns of patients with chronic low back pain and central sensitization: Insights from a machine learning method Xiaoping Zheng, Michiel Reneman, Bert Otten, Claudine Lamoth Abstract: Introduction: Chronic low back pain (CLBP) is the leading global cause of disability. Central sensitization (CS) is present in a subsample of patients with CLBP. Optimal physical activity (PA) is often recommended in the management of CLBP because it can reduce the risk of disability. However, the evidence of the relationship between PA intensity levels and CLBP is inconsistent, and the knowledge about the association with CS is limited. This study aimed to investigate PA patterns in patients with CLBP and low or high CS using an unsupervised machine learning approach. Methods: Forty-two patients were included (23 CLBP-, a CS Inventory score lower than 40; 19 CLBP+, 40-100). Patients wore a 3D accelerometer for about one week. For each patient, 4 days of data were used for analyses. Accelerometer data were corrected for gravity and the vector magnitude was calculated. For each group, a Hidden semi Markov Model (HSMM) was made to measure the temporal organization and transition of hidden states (PA intensity levels), based on accelerometer vector magnitude. Differences between CLBP- and CLBP+ in duration and occupation of hidden states were assessed with independent t-tests. The transition probability was assessed by Binomial-proportion test. The compositions of corresponding hidden states were assessed with Jensen–Shannon divergence (JSD). Results: The corresponding 5 hidden states of CLBP- and CLBP+ were similar, indicated by JSD. These states were defined as: rest (e.g., sleeping), sedentary (e.g., desk work), light activity (e.g., standing), light locomotion (e.g., slow walking), and moderate-vigorous activities (e.g., fast walking). Significant differences between 2 groups showed that CLBP+ exhibited higher duration and transition probability of active state (light activity, light locomotion, and moderate-vigorous states) and higher duration of inactive state (rest and sedentary states). Discussion: The significant differences in temporal organization and transition of PA levels may suggest that CLBP- and CLBP+ had different PA patterns. CLBP+ group exhibited a prolonged period of activity engagement (overactive) and then had a long period of rest. This PA pattern may suggest that CLBP+ had the distress-endures response pattern. 10:00 0 mins Analysis of abdominal aortic aneurysm thrombus using multispectral photoacoustic imaging Rick van Bergen, Camilo Cano, Min Wu, Marc van Sambeek, Richard Lopata Abstract: Background, Motivation and Objective The protocol for assessing abdominal aortic aneurysm (AAA) involves measuring the AAA diameter, which is found to introduce ambiguity. Consequently, several studies suggested that intraluminal thrombus (ILT) can be a better predictor for AAA rupture risk. Although the structure and composition of the ILT can contribute to a better assessment of AAA, it is challenging to image and differentiate the three main layers of an ILT, the luminal, medial, and abluminal layers. This difficulty stems from the low contrast between the layers. In this pilot study, we evaluate the feasibility of multispectral photoacoustic imaging (sPAI) on ILT to differentiate between these layers in an experimental setting. Statement of Contribution/Methods Different regions of two excised ILTs were scanned using two scanning schemes. The first is a transverse scan of the ILT where the different layers were imaged using equal optical fluence to explore the possibility of detecting layers based on PA spectral composition. The second is a luminal scan that mimics an intravascular approach. The results from the transverse measurement were spectrally unmixed using the piecewise convex multi-model endmember determination model. For the luminal experiment, an interface detection algorithm was developed. This algorithm detected the layers based on the bandwidth-limited PA signal from interfaces. The results were validated using scanning electron microscopy images of the samples. Results/Discussion In the transverse experiment, the tissue constituents found belonged to blood, fibrin, and the water/fibrin interface. With these endmembers and the corresponding proportion maps, it is possible to observe a correlation between the tissue constituent distributions and the ILT layers. For the luminal experiment, the unmixing results did not provide new insight due to limited optical fluence. However, the proposed interface detection algorithm successfully segmented the luminal layer of the thrombus. Moreover, the medial-abluminal interface was identified, yet a complete surface segmentation requires an improved penetration depth. To conclude, sPAI can be used to differentiate between the different layers of the ILT, although further steps are required, such as increasing penetration depth, to facilitate the translation from bench to bedside. 10:00 0 mins Current state of the art of automatic surgical instrument counting: A study protocol Anton Kooijmans, Maarten van der Elst, John van den Dobbelsteen Abstract: Background The operating room (OR) is a dynamic environment where many processes converge. They are complex and often time- and resource-constrained. This complexity compels OR personnel to perform many regular checks, resulting in an administrative burden that is counterproductive to the increasing demand for effectiveness and efficiency of an economic and safe patient care workflow. There is an urgent need for resource optimization and automatization of tasks to support humans as effectively as possible. Systems that automatically count surgical instruments before and after an operation have been developed, with the current state of the art using camera- and weight sensors. In this study, we investigate the reliability of one such system (Ancitrak, ATH Medical, Paris, France) in a real-world OR environment where it may be affected by variations in instrument condition during the procedure. Research question Primary question: What is the effect of instrument orientation and amount of soiling on correct classification? Secondary question: What is an acceptable rate of correct classification according to OR complex personnel? Methods Instrument trays used in TEPPs (Total ExtraPeritoneal Procedure) will be selected for this study. Complete ‘reference trays’ will be registered preoperatively in the Ancitrak system according to its standard procedure in which an image of the tray and its weight will be recorded. The detection performance will be evaluated for instruments arranged according to the reference tray as a control measurement. This will be compared to the performance for instruments that were used in actual procedures. Additionally, process time required for setting up the system, performing scans, etc. will be recorded to evaluate the impact on the workload of the staff. Furthermore, a questionnaire regarding an acceptable Ancitrak performance for implementation will be distributed among OR personnel. Analysis Primary outcome measure is the percentage of correct tray classification. Data will be analyzed using McNemar’s test. Complete tray data will be further analyzed using repeated measurements ANOVA. 10:00 0 mins Accurate pulse transit time estimation in low sampling frequency environments using parametric normalization Roel Montree, Elisabetta Peri, Reinder Haakma, Rik Vullings Abstract: Over the last years, Pulse Transit Time (PTT) has garnered a lot of interest to continuously and non-invasively determine blood pressure without using a cuff. The advancements in wearable technology have increased the accessibility to outpatient monitoring. However, because of computational limitations and a limited battery life, a reduction in the sampling frequency is desired. This comes at the cost of the accuracy of the measurements. To obtain relevant clinical information for cardiovascular health, accurate estimation of pulse parameters is vital. A new method has been developed to estimate the location of the foot of the pulse more accurately in a signal with a lower sampling frequency. This is achieved by matching with a template. The matching is done by reducing the calculated error between the pulse and the template, dependent on three parameters, while allowing for subsample matching. The method is kept low in complexity in consideration of the computational limits and battery consumption. This method is tested on a dataset recording the subject for the PTT between the heart as recorded by electrography (ECG) and plethysmography (PPG) recorded at the tip of the non-dominant hand index finger. Both are acquired at a sampling frequency of 500 Hertz. A comparison is made between the results at 500 Hertz and when the PPG signal is reduced to as low as 25 Hertz, both with and without the newly introduced method. Enabling more accurate measurement in low frequency environments allows the increase of data gathering outside of controlled environments, as well as continuous monitoring. Furthermore, large databases currently exist, recorded in low frequency, that can retroactively be used to characterize vital markers able to be found in theory and high frequency environments. 10:00 0 mins Comparison of 3D ultrasound and CTA in measuring abdominal aortic aneurysm characteristics – research plan Rianne E. van Rijswijk, Erik Groot Jebbink, Michel M.P.J. Reijnen Abstract: Background: An abdominal aortic aneurysm (AAA) is a dilation of the abdominal aorta that can lead to a lethal rupture. It can be treated with either open surgical repair or endovascular repair, in which an endograft is placed (EVAR). Studies show that patients with AAA sac shrinkage one-year after EVAR experience significantly better long-term outcomes than patients with a stable diameter [1,2]. In literature, AAA thrombus, total AAA volume, and lumen volume seem useful in predicting this remodeling of the aneurysm sac after EVAR [3]. These AAA characteristics are traditionally measured with contrast-enhanced computed tomography angiography (CTA), however, three-dimensional ultrasound (3D-US) is emerging as a novel, less harmful imaging alternative. With the use of SonoVue contrast agent, the AAA thrombus can also be distinguished from the flow-lumen on the US scans, similarly to CTA [4]. In ex vivo measurements, 3D-US has been shown to have clinically acceptable error rates with vessel diameter and volume measurement [5]. However, it is unclear whether this is also applicable to clinical measurements. Objective: The aim of this study is to compare preoperative AAA characteristics as measured by 1) 3D-US without contrast-enhancement, 2) 3D-US with contrast-enhancement, and 3) CTA. AAA characteristics to be investigated include AAA diameter, total AAA volume, lumen volume, thrombus thickness and thrombus volume. The secondary objective is to identify the added value of contrast-enhancement for 3D-US. Methods: Twenty patients with an AAA that are scheduled for EVAR will preoperatively undergo 3D-US with and without use of the Sonovue contrast agent, in addition to the standard CTA scan. Current status: The protocol is currently under ethical review. The aimed start of patient inclusion is December 1st, so by the end of January we hope to be able to present results of our first patients. Expected results: We expect to demonstrate that AAA characteristics can be as reliably measured with 3D-US, when compared to CTA. Also, we aim to show that contrast-enhanced 3D-US is feasible and establish the potential added value of contrast-enhancement in AAA patients. 10:00 0 mins Rigid and flexible multibody modelling of flexible instruments used in cervical cancer brachytherapy Robin Straathof, Jaap Meijaard, Sharline van Vliet-Pérez, Inger-Karine Kolkman-Deurloo, Remi Nout, Ben Heijmen, Linda Wauben, Jenny Dankelman, Nick van de Berg Abstract: ABSTRACT Purpose In brachytherapy (BT) for cervical cancer, a radioactive source is driven by cable through an intracavitary (IC) applicator placed in the vaginal and uterine cavity, and/or through interstitial (IS) catheters guided through the applicator and implanted in tissue. Accurate understanding of source cable and catheter behaviour is important due to the steep dose gradient in BT of up to 12% per mm. The purpose of this study is to develop and validate comprehensible computer models to simulate: (1) BT source paths, and (2) insertion forces of catheters in curved IC/IS applicator channels. These models can aid novel (3D-printed) BT applicator development and improve source path models used for treatment planning. Method Source cables or catheters were modelled as an interconnected series of rigid beam elements connected through torsional springs. For evaluating the source cable model, simulated paths of a Flexitron source cable (Elekta, Stockholm, Sweden) in CT/MR ring applicators (Elekta, diameters: Ø26, Ø30 and Ø34 mm, angles: 45° and 60°) were compared with manufacturer-specified source paths. For validating catheter models, simulated ProGuide 6F catheter with obturator (Elekta) insertion forces in S-shaped channels with varying design parameters (curvature, geometric torsion, and clearance) were compared with force measurements in dedicated 3D-printed templates. Results and discussion Median and maximum differences between simulated and manufacturer-specified dwell positions were 0.5 – 1.2 mm and <2.0 mm respectively in all but one ring applicator. Simulated catheter insertion force results were in close agreement with experimental results for all channel design parameters. Conclusion The developed models show promising results in predicting the behaviour of flexible instruments in BT applicators. Insights from these models can aid novel applicator design with improved motion and force transmission of BT instruments, and contribute to overall treatment precision. 10:00 0 mins Study of the locations and morphology of isolated knee focal cartilage defects using a statistical shape modeling approach Majid Mohammad Sadeghi, Erkan Aşık, Pieter Emans, Nazli Tümer, Gabrielle Tuijthof, Alex Roth Abstract: Focal knee resurfacing implants (FKRIs) are an emerging alternative treatment for knee focal cartilage injuries and are typically intended for the middle-aged population where biological cartilage procedures and total knee replacement are not proper options due to longevity concerns. The knowledge of the distribution of the exact location and size of the focal cartilage defects and the morphology of healthy cartilage at expected defect locations is a key requirement in designing off-the-shelf FKRIs. Current literature on cartilage defect characteristics is limited by the lack of detail and the subjective nature in which location and size are described, whereas little data is available on the local joint morphology. The objective of this work is to describe a statistical shape modelling (SSM)-based workflow for systematically and accurately describing the cartilage defect characteristics observed in patients targeted for FKRIs. Magnetic resonance imaging (MRI) scans of 50 patients treated with FKRIs were included in this study. Cartilage samples were segmented from these scans and were aligned into a common coordinate system using a registration algorithm. Cartilage defect zones were determined on MRIs and co-registered to the same coordinate system using the transformation found for the cartilage models transferred onto aligned cartilage samples using the corresponding transformation matrices obtained following their registration. The distribution of these defect properties, including their exact location in the coordinate system, and their sizes were determined using a dedicated algorithm. In parallel, a SSM of the cartilage was built from the cartilage samples. The mean and variations of morphological properties of the cartilage surface at the locations of the defects were found using the SSM model. The combined information of the distribution of the exact size and shape parameters of cartilage defects and the local cartilage morphology acquired from the SSM of the cartilage serve as important inputs for designing the resurfacing implants. This workflow proves to provide information on the isolated focal knee cartilage defects with the required level of detail currently missing in the literature for enabling FKRI design. This work may also be useful for guiding regenerative medicine or allografting approaches in cartilage repair. 10:00 0 mins A CT-derived intravascular ultrasound simulation framework for deep learning-based image segmentation Daniek van Aarle, Floor Fasen, Harold Schmeitz, Frits de Bruijn, Marc van Sambeek, Hans-Martin Schwab, Richard Lopata Abstract: Deep learning-based algorithms can extract patient-specific information on abdominal aortic aneurysms (AAAs) to improve current decision-making for clinical intervention. In this research we present a framework for the simulation of realistic two-dimensional intravascular ultrasound (IVUS) data based on Computed Tomography Angiography scans of the AAA. A segmentation network was trained on this synthetic dataset to prove the feasibility of deep learning-based image analysis on IVUS data. 10:00 0 mins Pelvin organ assessment in upright position pre and post pessary insertion Irina De Alba Alvarez, Anique Bellos-Grob, Frank Simonis, Frieda van den Noort Abstract: Introduction: Pelvic Organ Prolapse (POP) is the decent of pelvic organs due to the weakness of the pelvic floor muscles. Conservative treatment by means of pessary (silicone ring) treatment is first line of choice for POP patients. The purpose of a pessary is to lift the pelvic organs to reduce the physical POP complaints. Pessary placement, however, has a limited success rate, with reported successful pessary fitting of 41% [1]. The mechanical working mechanism (e.g. why does it stay in place or drop-out) is poorly understood. We previously established that POP is larger in upright as compared to supine position [2], [3] , which should be considered when assessing the orientation and location of the pessary. However, most studies are done in supine position without considering the effect of gravity on the pelvic floor. The aim of this project is to evaluate the effect of pessary insertion on the lowest point of the bladder and cervix, and the volume of the bladder with patients in upright position. Method: This prospective study included 15 POP patients, wearing a pessary for at least 3 months. Data was acquired with a tilting 0.25T MRI system. The PICS reference line was used for the measurements of organ location and volume [4], [5]. Statistically significant differences were determined by means of the Wilcoxon signed-rank test. Results: In upright position after pessary insertion, we found a decreased bladder volume (mean =2.30ml) as compared to no pessary in situ (mean =8.69ml) (p<0.01). The lowest point of the bladder was raised after pessary insertion from 30mm below the PICS line to 10mm below the PICS line (p<0.01). The lowest point of the cervix was raised after pessary insertion from 2mm above the PICS line to 21mm above the PICS line (p<0.01). Conclusion: When assessing the effect of pessary insertion in upright position, we can conclude that it lifts the bladder and uterus significantly. This study includes the first steps towards a full understanding of the true working mechanism of a pessary. 10:00 0 mins Multivariate recurrence plots for the analysis of Gait Joël Karel, Elena Heinze, Ralf Peeters, Pietro Bonizzi Abstract: Recurrence Plots (RP) are a tool for investigating and visualizing recurrent behaviour in dynamical systems. RP were originally designed for dealing with univariate signals. However, when measuring gait, typically three-axial accelerometers are used. In such a case it may be more convenient to create a multivariate RP instead of creating a separate RP for each of the accelerometer signals. A multivariate RP takes into account the redundant information of all given signals simultaneously and combines it in the same plot. In this respect, it is reasonable to think that a multivariate RP should be better equipped to capture the recurrent behaviour of a system than its univariate counterpart, when more signals are available from the system. To test this hypothesis, we analysed 20 seconds of gait of a healthy subject from the Long Term Movement Monitoring Database from Physionet, where there is a portion of gait of 13 seconds. For the univariate RP, the vertical of the accelerometer was used as this is the direction in which the gait is expected to reflect the most. For the multivariate RP, all three axes of the accelerometer were employed. This is illustrated below, where from left to right the univariate and the multivariate (using submatrix comparison) thresholded (epsilon=0.10) RPs are provided. It can be observed that the multivariate RP expresses the recurrent behavior more strongly as shown by the diagonal lines. At the same time, the noise in the plot decreases when using multivariate RPs. Hence, this explorative study gives indications that the use of multivariate RP may be beneficial for the analysis of 3D accelerometer signals, and for the joint analysis of multiple signals from the same system. 10:00 0 mins Power management in large-scale multichannel stimulator systems Francesc Varkevisser, Tiago L. Costa, Wouter A. Serdijn Abstract: The development of neuroprosthetic bioelectronic devices requires implantable stimulator systems with hundreds to thousands of output channels [1]. Typically, these devices are powered wirelessly, which leads to limited available power. Therefore, power efficiency optimization is crucial to facilitate as many output channels as possible. Current-mode electrical stimulation is favored over voltage-mode stimulation because of its inherent control over the charge injected into the tissue, which is important for safety. However, the voltage drop over the current driver leads to power losses, especially in multichannel stimulators. An adaptive voltage supply can reduce overhead losses and increase power efficiency [2]. Still, due to area constraints, providing channel-specific adaptive voltage supplies in large-scale multichannel systems remains challenging. Moreover, conventional power management implementations use multiple conversion stages, which leads to cascaded losses and low total conversion efficiency [3]. This work proposes a parallel, adaptive ac/dc power management strategy for generating channel-specific adaptive voltage supplies. A phase-controlled adaptive ac/dc converter is designed that converts the incoming power signal to the appropriate voltage level in a single stage, preventing cascaded losses. Multiple converters are employed in parallel to generate channel-specific voltage levels to increase the power efficiency at each output channel. The proposed implementation requires less area and offers better scalability than the alternative approach of switched-capacitor dc/dc converters. The power management design will be combined with current-mode stimulator circuits to develop a large-scale, power-efficient multichannel system for neuroprosthetic applications. [1] Fernández, E., Alfaro, A., & González-López, P. (2020). Toward Long-Term Communication With the Brain in the Blind by Intracortical Stimulation: Challenges and Future Prospects. Frontiers in Neuroscience, 14(August). [2] Lee, H.-M., Park, H., Ghovanloo, M. (2013). A Power-Efficient Wireless System With Adaptive Supply Control for Deep Brain Stimulation. IEEE Journal of Solid-State Circuits, 48(9), 2203–2216. [3] Kim, C., Ha, S., Park, J., Akinin, A., Mercier, P. P., & Cauwenberghs, G. (2017). A 144-MHz Fully Integrated Resonant Regulating Rectifier with Hybrid Pulse Modulation for mm-Sized Implants. IEEE Journal of Solid-State Circuits, 52(11), 3043–3055. 10:00 0 mins Intravascular ultrasound to obtain geometry and local wall thickness of the abdominal aorta Floor Fasen, Daniek van Aarle, Marc van Sambeek, Arjen van der Horst, Richard Lopata Abstract: An abdominal aortic aneurysm (AAA) is a localized widening of the abdominal aorta. AAA can lead to death because of rupture of the wall and therefore a clinical intervention is performed when the diameter of the aorta is too large. However, a better patient-specific marker is needed. The mean wall thickness is an indicator for rupture risk [1], which varies significantly within and between patients [2]. So far, regional thickness has not been used in previous rupture risk analysis studies, since it cannot be measured with CT, MRI and non-invasive ultrasound (US). In this study, a method using intravascular ultrasound (IVUS) is proposed. Four setups were used to acquire IVUS (Philips PV Visions 0.035) images of aortas. First, two poly(vinyl alcohol) phantoms were created with a given wall thickness. The phantoms were placed in a mock circulation set-up to mimic in vivo hemodynamic conditions. Second, seven static porcine aortas were imaged using IVUS and μCT, where the μCT images serve as ground truth. Third, 347 IVUS images of AAAs were simulated based on CTA images. Fourth, three AAA patients were imaged during surgical intervention. For the porcine aortas, two automatic segmentation methods were adopted, optimized and compared to segment the lumen and outer wall: a gradient-based and an active contours method. Finally, the geometry and wall thickness were obtained. The segmentation quality was investigated using the Hausdorff distance (HD) and the similarity index (SI). The segmentation quality of the inner wall (HD=0.9 mm, SI=0.96) is better than the outer wall (HD=1.3 mm, SI=0.95). The active contours method (HD=0.8 mm, SI=0.97) outperforms the gradient-based method (HD=1.4 mm, SI=0.94). In this study we showed that using IVUS, the geometry and wall thickness of the aorta can be obtained accurately. This method can be used to improve AAA rupture risk assessment in a minimally invasive manner. In future studies, the phantom, simulated and patient data will be explored using the developed conventional methods and a convolutional neural network. For large aneurysms, the wall is not always in the field of view. Hence, a combined intravascular-extravascular approach will be considered. [1] Di Martino, E. S., et al. Journal of Vascular Surgery (2006). [2] Raghavan, M. L., et al. Journal of biomechanics (2006). 10:00 0 mins Imaging and mechanical characterisation of human blood clot analogues with different compositions and levels of contraction Rachel Cahalane, Janneke Cruts, Ahlam Rachid, Kim van Gaalen, Heleen van Beusekom, Moniek de Maat, Marcel Dijkshoorn, Nikki Boodt, Aad van der Lugt, Frank Gijsen Abstract: Background Acute ischemic stroke is caused by a thrombus that blocks an intracranial artery. The efficiency of a thrombus removal procedure is thought to be influenced by the thrombus mechanical properties, which are known to be affected by both the thrombus composition and contraction. Computed tomography (CT) could be a suitable modality to assess thrombus properties prior to the intervention.[1] We examine the imaging characteristics and mechanical properties of clot analogues with different compositions and levels of contraction. Methods Clot analogues were made from citrated whole blood of six healthy human donors. Clots were made with five different red blood cell (RBC) volumes: 0, 25, 77, 94 and 99%, which span the range of compositions from ex vivo thrombi. Also, clots were made with three different platelet concentrations: 30*103, 90*103, and 270*103 platelets/μL, which result in a low, medium and high level of contraction, respectively. Clinical CT imaging was performed to measure the density. Perviousness, which reflects the clot’s permeability, was quantified by measuring the density after the administration of a contrast agent. Unconfined compression tests were performed and the high strain stiffness was obtained from the slope of the stress-strain curves at 75-80% strain. Results The compressive stiffness was highest for the highly contracted clots and lowest for the low contracted clots. Significant differences in stiffness were also observed between RBC volume groups within the high and medium contracted clot groups. For the first time we demonstrated that higher levels of contraction increase the CT density values. As expected, the density values were higher for increasing RBC volume. Highly contracted clots tended to have higher perviousness values compared with medium and low contracted clots. There was only a significant difference in perviousness between the RBC volume groups within the highly contracted clot group. Conclusion Clot compressive stiffness and CT density are dependent on the level of contraction and RBC content. The perviousness appears to increase with increasing level of contraction, but is only affected by RBC content in highly contracted clots. 10:00 0 mins Moving out of the Lab: Can we estimate Ground Reaction Forces in Running with 3 Intertial Measurement Units? Bouke Scheltinga, Jaap Buurke, Jasper Reenalda Abstract: Running is a sport with a high injury incidence. Monitoring biomechanical load could help in understanding the development of these injuries. Ground reaction force (GRF) can be seen as an important measure to quantify biomechanical load during running [1]. However, GRF measurement is restricted to the lab oratory. Artificial neural networks (ANNs) are capable to model complex relations and thus could be used to predict GRF from inertial measurement units (IMUs) [2]. This prediction would be a step towards the quantification of biomechanical load outside the lab. The goal of this abstract is to show the possibilities of a generic ANN to estimate 3-dimensional GRFs using three IMUs. Methods: 11 experienced heel strike runners (4F, 7M; 30.6y ± 8.3y, 1.80m ± 0.11m, 74.2kg ± 17.4kg ) ran 9 trials on a force-instrumented treadmill at combination of three velocities (10, 12 and 14km/h) and three stride rates (preferred, -10% of preferred and +10% of preferred). Subjects were instrumented with IMUs (240Hz) mounted at both proximal tibias and pelvis. Using leave-one-subject-out cross validation, for each subject a model was created while data from the other subjects where used for training and validation of the model. Two layer ANNs (100 neuron s each) were then trained with the 3D gravity subtracted acceleration in the global frame as input to fit the 3D GRF. Performance of the models was analysed with the root mean squared error (RMSE) in body weight and as percentage of the range and Pearson’s correlation coefficient. Results: The ANN modelled GRFs with high accuracy for the forces in the sagittal plane with a RMSE ≤ 7% for the vertical direction and a RMSE ≤ 10% for the anterior-posterior direction for 8 out of the 11 subjects. Over all subjects, Pearson’s correlation coefficients of 0.96, 0.86 and 0.46 were achieved for respectively the vertical, anterior-posterior and medio-lateral direction. Conclusion: ANNs can be used to predict GRFs at a good accuracy for the vertical and anterior-posterior direction, but not yet in the medio-lateral direction. The variability in the performance for the different subjects could be explained by the variability in running gait between subjects. [1] Verheul, J., Nedergaard, N. J., Vanrenterghem, J., & Robinson, M. A. (2020). Measuring biomechanical loads in team sports–from lab to field. Science and Medicine in Football, 4(3), 246–252. https://doi.org/10.1080/24733938.2019.1709654 [2] Wouda, F. J., Giuberti, M., Bellusci, G., Maartens, E., Reenalda, J., van Beijnum, B.-J. F., & Veltink, P. H. (2018). Estimation of Vertical Ground Reaction Forces and Sagittal Knee Kinematics During Running Using Three Inertial Sensors. Frontiers in Physiology, 9, 1–14. https://doi.org/10.3389/fphys.2018.00218 10:00 0 mins Continuous implicit neural representations for personalised 3D vascular models Dieuwertje Alblas, Christoph Brune, Kak Khee Yeung, Jelmer Wolterink Abstract: Personalised 3D vascular models are valuable for diagnosis and treatment planning in patients with cardiovascular diseases. Acquiring explicit representations of these models from image data is time-consuming, and often requires pre-processing for downstream tasks. We propose to represent these surfaces implicitly by the zero levelset of their signed distance function in an implicit neural representation. For training, this network only requires a set of points on the vessel’s surface. The network represents the surface at any resolution with high accuracy. We briefly describe the method and its benefits. The signed distance function (SDF) of a surface M is defined as: SDF_M (x)= {█(-d(x, M)@0@d(x, M) )┤ ■(x inside M @x on M @x outside M) Where d(x, M) is the distance to the surface. The SDF is a continuous function that can be evaluated at any x in R^3. We use a neural network f(x;θ) to approximate the SDF [1, 2]. Similar to the SDF, the network takes in coordinates x and outputs the local SDF value. For training, we used a set of points X⊂R^3 on the vessel’s surface. We assessed the robustness and accuracy of this method by reconstructing 3D mesh models of the abdominal aorta. We can reconstruct the vessels with a Dice Similarity Coefficient of 0.94 with as few as 200 points. Moreover, we reconstructed the SDFs of layered structures in the vessel wall with a single neural network. The neural network captures the nested topology of the shapes, resulting in zero unwanted extrusions of inner structures, in contrast to representing each shape with a separate neural network. Lastly, smoothly connecting multiple vessels into a vascular tree is cumbersome for mesh models. In our approach, multiple vessels can be easily combined into a smooth vascular tree using a smooth minimum operation on their SDFs. We demonstrated a method that reconstructs smooth, watertight vessels, based on a set of points. This can be readily adapted into an annotation pipeline to acquire personalized 3D vascular models from image data. Smooth blending of these vessels makes the obtained vascular models directly suitable for downstream tasks, e.g. computational fluid dynamics. 10:00 0 mins Removal of electrocardiographic interference, noise, and artifacts from diaghragm electromyography Gabriela Grońska, Elisabetta Peri, Xi Long, Hans van Dijk, Massimo Mischi Abstract: Introduction: Diaphragmatic electromyography (dEMG) is a non-invasive method to monitor respiratory activity. However, the quality of dEMG is often compromised by the presence of multiple disturbances, such as cardiac electrical activity, noise, and motion artifacts. This impacts the clinical uptake of dEMG to monitor respiration. A recent paper from our group proposed an algorithm based on singular value decomposition (SVD) to remove cardiac interference, showing its superiority with respect to other strategies proposed in the literature [1]. Nevertheless, the method is not satisfactory in presence of motion artifacts. In this work, we aim at improving the performance of the SVD-based algorithm to remove both cardiac interference and motion artifacts. Methods: The original algorithm used signal-to-noise ratio as a criterium to select singular values corresponding with ECG contamination. In the current work, we investigate a new index to reject singular values which map components corresponding not only with ECG contamination but also with motion artifacts. The index was created based on the frequency domain range of ECG and motion artifacts (mainly 5-30Hz). Power spectral density (PSD) was obtained for each singular value, and the ratio between PSD5-30Hz and PSD30-250Hz was calculated as an index. Components with the highest indexes (outside 95% of the distribution) were rejected. A synthetic dataset was used to compare the performance of the algorithm with the new index to the original algorithm. Muscular activity of the biceps from nine healthy volunteers was combined with ECG only (dataset A) and with ECG mixed with motion artifacts (dataset B). The error in the EMG reconstruction was assessed as the root mean squared error (RMSE) in the time domain and mean-frequency difference (MFD) in the frequency domain, and it is reported as median (interquartile values). Results and Discussion: The proposed method achieved an RMSE of 9.9(1.8)%, and an MFD of 0.31(0.02)Hz for dataset A. The performance on dataset B showed an RMSE of 28.2(2.3)% and an MDF was 0.59(0.56)Hz. Results are significantly better than what was achieved with the original algorithm in time and frequency domains for both datasets (p-value<0.03). In future work, the proposed algorithm will be tested on real data. References [1] Peri E., et al. Sensors (2021); 21, 573 10:00 0 mins Finite element study on the effect of varying the interference fit and coefficient of friction on the primary fixation of a cementless peek tibial component Corine Post, Thom Bitter, Inger van Langen, Adam Briscoe, Nico Verdonschot, Dennis Janssen Abstract: Introduction: Cementless polyetheretherketone (PEEK-OPTIMATM) knee components are of interest due to its stiffness properties being more similar to the properties of bone than titanium. This may potentially avoid stress-shielding. The primary fixation of cementless knee components are crucial for the long-term fixation, with the micromotions and gaps as important quantifications. The interference fit and coefficient of friction both influence the primary fixation and therefore need to be researched in cementless PEEK tibial components. In this finite element (FE) study, the effect of varying the interference fit and coefficient of friction on the micromotions of a cementless PEEK and titanium tibial component was investigated. Methods: Nine FE models were created of the tibia and PEEK or titanium tibial tray including polyethylene insert (5L, 4R). Gait and squat activities were simulated. Two variations of interference fit (500, 750 µm), defined through the contact algorithm and applied at the bone-implant interface, and two variations of coefficient of friction (0.5, 1.5) were investigated. To quantify the primary fixation, the 95th percentile of the resulting shearing micromotions, resulting normal gaps and resultant (vectorial) micromotions were analyzed for each FE simulation. Results: Both the interference fit and coefficient of friction did not have a large impact on the shearing micromotions. The high interference fit resulted in slightly larger normal gaps and resultant micromotions. Additionally, the micromotion and gap values of the cementless PEEK tibial component were similar to the micromotion and gap values of the more traditional cementless titanium tibial component. Conclusion: This study showed that the shearing micromotions of a cementless tibial component has minor effects caused by the in this study researched variations of interference fit, coefficient of friction and implant material. The interference fit has small effects on the normal gaps and resultant micromotions. The current FE simulations give understanding of the influence of interference fit and coefficient of friction on the micromotions and gaps. This may be further assessed in a larger population of models to analyze the sensitivity of patient related characteristics on micromotion outliers. Acknowledgements: PEEK-OPTIMATM is a trademark of Invibio Ltd. Implant geometry was supplied by Maxx Orthopaedics Inc. 10:00 0 mins Defining the surgical target of optimal implant positioning in robotic-assisted total knee arthroplasty Periklis Tzanetis, Kevin De Souza, René Fluit, Seonaid Robertson, Bart Koopman, Nico Verdonschot Abstract: One of the challenges in robotic-assisted total knee arthroplasty (TKA) is to define the optimal implant position in order to achieve restoration of the premorbid joint state. The aim of this study was first to explore differences in tibiofemoral kinematics and ligament strains between arthritic knees and their premorbid reconstruction, and subsequently, to optimize the position of the femoral component in the implanted knee, thereby reproducing the premorbid kinematics as closely as possible. A template musculoskeletal cadaveric knee model was developed using a previously established methodology [1]. To personalize the model, a computed tomography (CT) image of a single patient with severe knee osteoarthritis was segmented using a CT-based statistical shape model (SSM) trained to identify the arthritic and premorbid femoral and tibial bone surfaces [2]; and a magnetic resonance imaging-based SSM was used to estimate the cartilage thickness from the segmented bones. A cruciate-retaining implant was mechanically aligned to the reconstructed premorbid knee. The femoral component’s position was varied from −6 to 6 mm for medial-lateral, anterior-posterior, and proximal-distal translations, −3° to 3° for flexion-extension, and −6° to 6° for varus-valgus and internal-external rotations, using the covariance matrix adaptation evolution strategy. Compared to the premorbid model, the mechanically implanted model predicted tibiofemoral kinematics with an average root-mean-square-deviation (RMSD) of 2.53±1.24 mm and 2.45±0.98°. We found an average RMSD of 0.95±0.31 mm for translations and 1.35±0.01° for rotations between the premorbid and optimally aligned TKA model. The ligament strains reported deviations less than 1% over the extension range from 60° to 0°, except for the deep medial collateral ligament. We demonstrated the feasibility of closely reproducing premorbid kinematics in robotic-assisted TKA by optimizing the position of the femoral component with respect to mechanical alignment. Different surgical targets for robotic-assisted surgery can be further discussed, such as maximum joint stability in the mid-flexion range, balanced collateral ligaments, or reproducing the premorbid ligament strains. 1. Marra et al, J Biomech Eng, 137:020904, 2015. 2. Bowes et al, Ann Rheum Dis, 80:502-508, 2021. 10:00 0 mins The impact of posterior tibial slope on knee biomechanics as predictor of flexion contracture in total knee arthroplasty Dionne Brandsma, Periklis Tzanetis, René Fluit, Bart Koopman, Nico Verdonschot Abstract: Pre-operative knee flexion contracture in patients undergoing total knee arthroplasty (TKA) is a risk factor for post-operative flexion contracture associated with undesirable surgical outcomes such as restricted range of motion and knee instability [1]. Osteophyte formation [2] and changes in the posterior tibial slope (PTS) [3] in osteoarthritic knees may contribute to flexion contracture. However, the effects of these morphological changes are not well understood. The aim of this study is to investigate the effect of PTS on the tibiofemoral joint kinematics and ligament and muscle forces using a musculoskeletal modeling approach. A patient-specific musculoskeletal knee model will be developed using a previously established framework [4] to represent the pre-operative knee comprising the osteophytic femoral and tibial bones. The PTS will be the only varying parameter in the model, while all other variables, such as the muscle and ligament attachment sites and their slack lengths will remain unchanged throughout the entire study. The PTS in the pre-operative knee will be determined based on anatomical landmarks located at the proximal tibia. The PTS will be varied from -9° up to 3° with increments of 3° compared to the baseline by rotating the tibial plateau in the sagittal plane of the patient-specific anatomical frame. We will capture the effect of PTS on the tibiofemoral joint kinematics, ligament and muscle forces during an unloaded knee extension simulation from 60° to 0°. Differences in the simulated outcomes will be quantified using the root-mean-square deviation and Pearson’s correlation coefficient. This study aims to provide a better understanding of the impact of changes in bone morphology on the pre-operative flexion contracture in osteoarthritic knees using patient-specific musculoskeletal modeling. The outcomes of this study could be valuable in the pre-operative planning of TKA and may assist surgeons to optimize their surgical decision-making approach to correcting flexion contracture. 1. Su, J Bone Joint Surg Br, 94:112–5, 2012. 2. Leie et al, J Orthop, 19:76–83, 2020. 3. Mochizuki et al, PLoS One, 13:e0202488, 2018. 4. Marra et al, J Biomech Eng, 137:020904, 2015. 10:00 0 mins Patient-specific 3D virtual surgical planning for sacroiliac joint fusion Nick Kampkuiper, Jorm Nellensteijn, Edsko Hekman, Maaike Koenrades, Femke Schröder Abstract: Low back pain is the leading cause of years lived with disability worldwide. In 15-25% of the patients this is caused by sacroiliac (SI) joint dysfunction. When conservative therapy is insufficient, it can be treated with sacroiliac joint fusion (SIJF). This involves stabilisation of the joint with three cannulated implants. Disadvantages of the procedure are radiation exposure, implant loosening, anatomical variations, and the use of solely 2D fluoroscopic guidance. Therefore, it can be challenging to place implants in a stable patient-specific configuration without damaging critical structures. Virtual Surgical Planning (VSP) might be the solution. This study aimed to introduce VSP in SIJF, evaluate the accuracy of the surgical procedure, and determine the surgeon's learning curve. The VSP consists of virtually inserted implants inside a preoperative CT scan and virtual fluoroscopic images of the pelvis with inserted implants that are used during the procedure. In a quantitative accuracy evaluation, three measures were used to assess the 3D deviations of implant placement. In the learning curve analyses the mean 3D deviations per procedure were assessed for 24 consecutive clinical cases. Additionally, the added value based on the surgeon's perspective was discussed. VSP for SIJF was successfully introduced. The results showed relatively low 3D deviations between the implants of approximately 4.8 mm between the apexes, an angular deviation of 4.1°, and 3.9 mm between the entry points, indicating that the surgeon could reproduce the VSP adequately. In the learning curve analysis, however, no trend was visible. The surgeon considers VSP as a valuable addition to SIJF since the surgeon felt better prepared, was more confident, and could perform the surgery more safely. Besides, the surgeon thought that due to VSP longer implants could be used and the duration of the surgery decreased. This study developed and implemented VSP for SIJF into clinical practice. Since in literature no accuracy measures were available for SIJF a comparison was not possible. It was expected that the accuracy would increase when the surgeon becomes more proficient in using VSP. A randomised control trial is recommended to assess the added value of VSP compared to the conventional procedure. 10:00 0 mins Development of a perfusion phantom for DCE-BCT Liselot Goris, Srirang Manohar, Ioannis Sechopoulos Abstract: Purpose Dynamic contrast-enhanced dedicated breast CT (DCE-bCT) is a novel breast imaging technique in which functional information of the breast is captured by evaluating the contrast agent uptake and washout over time.1 This information could be useful for treatment planning, response monitoring, and prediction in patients with breast cancer.1 This research aims to validate and evaluate the quantitative accuracy of DCE-bCT imaging by performing a physical phantom study. Methods First, a method for real-time iodine contrast concentration monitoring in a breast phantom using optical absorption spectroscopy will be developed. The setup will include a pump for water and contrast bolus input, tubing, a light source that emits light through the cross-section of the tubing and a detector to capture the light spectrum. Additionally, a phantom with the ability to simulate the perfusion of fibroglandular and tumour tissue will be developed. With DCE-bCT, time-intensity curves of volumes of interest in the phantom will be obtained. The absorption spectroscopy setup will be extended to be used with the breast perfusion phantom in DCE-bCT. Fibre optic cables will be incorporated to quantify contrast flow in the phantom’s tissue compartments during DCE-bCT measurements without having the light source and detector inside the field of view. In the next step, the model will be expanded to simulate the perfusion of tumours with varying morphologies. The final step will be to use the phantom to evaluate the quantitative performance of different imaging modalities. Expected results The absorption spectra for different iodine contrast concentrations and the relation of the peak absorption to the iodine concentration will be evaluated. This relationship will be used to capture the contrast concentration over time with spectroscopy. This technique will be used to validate the contrast concentration over time estimated with DCE-bCT by comparing the results. Conclusion In this work, the plans for a phantom study to validate and evaluate quantitative DCE-bCT imaging are presented. The phantom may provide insights into other multi-modality breast perfusion imaging approaches too, such as DCE-MRI and DCE-DBT. 10:00 0 mins Modeling of the hemodynamic response using functional ultrasound Sofia-Eirini Kotti, Aybüke Erol, Borbála Hunyadi Abstract: Functional ultrasound (fUS) is an emerging technique that images cerebral blood volume (CBV) changes with high sensitivity, without the use of contrast agents. In other words, fUS detects blood flow, or the number of moving red blood cells, in the voxels. These CBV changes are caused by the increased metabolic demand of active tissue and, thus, reflect neuronal activity in the corresponding brain area. The main advantages of this technique are that it can image the whole depth of the brain with high spatial (50-500um) and temporal resolution (10-100ms), and that it constitutes a potentially portable solution, as opposed to functional magnetic resonance imaging, the currently predominant modality in functional brain imaging. The fundamental challenge that comes with this technique is that it only provides an indirect measure of brain activity through the neurovascular coupling (NVC), which is the link between local neural activity and the resulting changes in the cerebral blood flow. This is a system with dynamic and non-linear characteristics which are only partially known. Moreover, besides the activity of interest, fUS records a mixture of other ongoing brain activity, physiological artifacts and noise. The goal of this research is estimate the brain’s hemodynamic response function (HRF), which is the mathematical representation of the NVC. The HRF is commonly described in literature using a linear time invariant system. In this work, we assume a nonlinear time invariant model that describes the relationship between the brain’s response to stimuli and the measured voxel time courses, and we use higher order kernels to characterize this model. Our results on fUS data obtained from mice reveal that including nonlinearities in the HRF achieves a significantly more precise modelling of the fUS signal compared to the linear system assumption under specific stimulus conditions. 10:00 0 mins Combined photoacoustic imaging and fiber Bragg grating sensors-based thermometry to monitor thermal ablation procedures Leonardo Bianchi, Hindrik Kruit, Paola Saccomandi, Srirang Manohar Abstract: Minimally invasive thermal therapies have gained rising acceptance as alternatives to traditional cancer surgery. Thermal techniques consist of the therapeutic application of thermal energy to induce the desired injury in the tumorous region. The provoked thermal damage in tissue is both dependent on the temperature reached and the persistence of the tissue medium at that temperature. Tissue temperatures in the hyperthermia range should be maintained for 30 – 60 min to obtain irreversible damage, whereas higher temperatures (≥ 60 °C), typically reached during thermal ablation (TA) procedures, are related to nearly instantaneous irreversible injury [1]. The main advantages of TA techniques are reduced pain, lower trauma and shorter recovery stay compared to surgery, as well as a lower requirement for radio/chemotherapy. Hence, TA methods, such as percutaneous radiofrequency ablation, are already commonly employed for the treatment of unresectable liver lesions. Nevertheless, relapses after this procedure are frequently observed due to incomplete ablation. Indeed, the effectiveness of TA procedures is still limited due to the lack of a real-time monitoring system [2]. This leads to the difficulty in delivering the proper thermal dose to completely eradicate the tumor mass, without endangering the integrity of the surrounding tissue. In this study, we devise a combined approach based on photoacoustic imaging and quasi-distributed temperature sensors, to assess and monitor the thermal effect on ex vivo hepatic tissue undergoing TA. The temperature sensors are based on fiber Bragg grating (FBG) etched in optical fibers. Photoacoustic signals arising from two different wavelengths (i.e., 760 nm and 970 nm) are acquired to allow for image guidance of the ablation device, identification of the ablation region, along with estimation of the ablation zone progression during the treatment. The FBG thermometers provide real-time feedback on the tissue temperature variation. Hence, exploiting the multiplexing and multipoint measurements capability of these sensors, spatially resolved two-dimensional thermal maps within the ablated tissue are reconstructed, during the procedure, through the analysis of the optical spectra registered by the FBG array. This preliminary study sets the basis for the establishment of a multi-sensor platform for fine-tuning TA procedures for the eradication of tumor lesions. 10:00 0 mins Label-free higher harmonic generation microscopy for survival verification of 3D lung tissue cultures Yuanyuan Ma, Sophie Schilder, Bonnie Plug, Laura van Huizen, Jan Willem Duitman, Teodora Radonic, Marie Groot Abstract: Pulmonary fibrosis is a fatal disease causing severely disrupted gas exchange and a gradual decline in lung function over time. To further understand the pathogenesis of pulmonary fibrosis, 3D lung tissue cultures (3D-LTCs) are excellent model systems to mimic the 3D lung microenvironment structure and dynamics in vivo. So far, real-time fluorescence imaging microscopy has been used to this end. However, fluorescence microscopy requires labelling of the tissue, causing unwanted side effects such as shortening the lifetime of tissue due to reactive oxygen species. Label-free higher harmonic generation microscopy (HHGM) has the advantages of a low and safe laser power and fast imaging speed, and therefore is expected to overcome these limitations. In this study we performed a primary viability test. Mouse 3D-LTCs were cultured in a well maintained environment inside an incubator attached to a microscope and imaged with HHGM for seven days, then the HHGM images were compared with immunohistochemical and histological staining. The HHGM images visualized macrophages, collagen fibers, alveoli, bigger airways, blood vessels and the pleura. The 3D-LTCs stayed viable for about five days after which the proliferation of cells in the 3D-LTCs dramatically decreased and apoptosis increased. This viability test shows that HHGM is a promising tool to study 3D-LTCs. Next, we will image the dynamics between cells and connective tissue in the LTCs, mimic the pulmonary fibrosis process and test therapies, with the ultimate aim of improving therapy of patients suffering from pulmonary fibrosis. 10:00 0 mins Sitting behaviour detection to reduce the risk of non-communicable disease In office workers – A study protocol Linda Ong, Ming Cao, G.J (Bart) Verkerke, Claudine J.C Lamoth, Elisabeth Wilhelm Abstract: Musculoskeletal disorders (MSDs) and metabolic syndrome (MetS) are two common health conditions in office workers. According to WHO, 1.7 billion people worldwide experience MSDs [1]. Around a quarter of cases occur in Europe [2] and MSDs frequently are related to the neck, upper, and lower back regions. Furthermore, MetS, which is known as a risk factor of developing chronic diseases, such as diabetes type II, and cardiovascular diseases, occurs up to 32.6% in office workers in some European countries [3]. Currently, support for the prevention of these two health issues is limited. People only seek treatment once the symptoms are perceived and they are limited in daily functioning. Unfortunately, body function limitation indicates MSDs and MetS might be already in the chronic phase, which will be more difficult to treat and reverse the condition. Therefore, the prevention of developing MSDs and MetS is necessary. One of the main risk factors for developing MSDs and MetS is sedentary behaviour during performing work that requires static and prolonged sitting. Even sitting upright can cause serious health issues if one keeps a static posture for a prolonged time. Therefore, the main aim of the study is to detect patterns in sitting behaviour, including different postures, sitting transition in postures while sitting, and duration of static postures. Existing sitting posture detection algorithms were based on pressure mat data in a controlled environment [4]. We will develop a posture detection algorithm based on real-life environment measurements. The study will involve 82 participants that spend most of their working hours (4-8 hours) in their working space sitting. The sitting monitoring is done for 5 working days. 80% of the data will be used for training, 10% for development, and 10% for testing. Sitting posture classification will be developed based on a novel machine learning approach while sitting transition and duration will be identified by thresholding the result from posture classification. To be useful for health advice, the system should classify common sitting postures as upright, slouch, slump, lean left, right, and backward with accuracy and f1-score of at least 85% and 80% respectively. The study has been running since mid-July 2022 and per October 2022, 15 participants have been following the whole measurement procedure. 10:00 0 mins Outlier resistant tensor decompositions for denoising multiplicative noise Metin Calis, Alle-Jan van der Veen, Massimo Mischi, Bori Hunyadi Abstract: In most ultrasound despeckling applications, the noise is considered multiplicative and Rayleigh shaped. Commonly the logarithm is taken, transforming the multiplicative noise into additive noise. According to [1], the logarithm of the noise follows Fisher-Tippet distribution, which can be approximated as white gaussian noise with outliers. In our previous work [2], we assumed that the noise was approximately white gaussian and recovered the clean ultrasound data as a low-rank approximation of the noisy observations via multilinear singular value decomposition (MLSVD). In a simulation study, we show that the true rank of the underlying signal could be approximated most of the time correctly. On the in-vivo recording of 6 patients, we showed that truncated MLSVD led to a better classification of cancerous voxels using dispersion and perfusion features. However, MLSVD is not optimal in the least-squared sense and is not robust against outliers. The estimate of the underlying low-rank tensor can be improved with prior information about the noise statistics. Therefore, in this work, we improve our signal model, assuming that the received signal is the summation of a low-rank tensor, white gaussian noise, and sparse outliers. In order to recover the clean contrast-enhanced ultrasound data based on this signal model, tensor stable component pursuit or non-convex outliner resistant tensor decompositions techniques can be used. We will compare the performance of these approaches with MLSVD and show the added benefit of the prior information about the noise statistics. [1] Michailovich, O. V., & Tannenbaum, A. (2006). Despeckling of medical ultrasound images. IEEE transactions on ultrasonics, ferroelectrics, and frequency control, 53(1), 64–78. https://doi.org/10.1109/tuffc.2006.1588392 [2] Calis, M., Mischi, M., van der Veen, A. J., & Hunyadi, B. (2022). Denoising of Dynamic Contrast-enhanced Ultrasound Sequences: A Multilinear Approach. In BIOSIGNALS (pp. 192-199). 10:00 0 mins Quantitative analysis of vaginal and uterine geometry for patient-tailored brachytherapy applicator development Sharline van Vliet - Perez, Tim Crone, Ruben van den Broek, Robin Straathof, Linda Wauben, Nick van de Berg, Jenny Dankelman, Ben Heijmen, Inger-Karine Kolkman - Deurloo, Remi Nout Abstract: Objective The currently used applicators for locally advanced cervical cancer (LACC) brachytherapy (BT) do not optimally align with individual patient anatomy and have fixed catheter positions and angles limiting treatment plan conformity. Patient-tailored applicators with optimally planned catheter channels could be the solution for this problem. To develop patient-tailored applicators, and to create a representative data set for safety tests and quality management, data on the dimensions of the distended vaginal and uterine shape is required. Little is known about the distended geometry of LACC patients after radiochemotherapy. Therefore, the aim of this study is to evaluate the vaginal and uterine geometry distended with ultrasound gel in LACC patients. Method In total 20 pre-BT MRIs of LACC patients with ultrasound gel to distend the vaginal cavity were included. The lengths and angles of the mid vagina (P0-P1), upper vagina (P1-P2), anterior cervix (P2a-P2), posterior cervix (P2-P2b), endocervical canal (P2-P3), and uterine body (P3-P4) were determined. Furthermore, the left-right (LR) widths of the vaginal cavity at height of the inferior pubic point (P0), internal ureteral ostium (P1), and external ostium (P2) were evaluated. In addition, frequency distribution maps of the vaginal cavity were made based on 3D models of the delineated MRIs. Results The lengths and angles (mean ± SD) between the vaginal and uterine segments and the LR widths were calculated. The largest variation was seen in the region P2-P4 as indicated by the length of P2-P3 (33 ± 11 mm), the width at P2 (34 ± 16 mm), the angle in P3-P4 in relation to the x-axis (123 ± 53°), and the frequency distribution maps. The mean volume of the 3D models was 48 (range: 12-69) cm3. Conclusion In this study the distended vaginal and uterine geometry was evaluated in LACC patients. The large inter-patient variation in length, angle, and width of the vaginal cavity suggests a patient-tailored applicator design, especially in the top part of the vaginal cavity. The evaluated parameters and frequency distribution maps are now available and can be used as design parameters and for safety and quality assessment of the patient-tailored applicators for LACC BT. 10:00 0 mins Embodiment of virtual feet modulates motor performance in a stepping task Alex van den Berg, Bart de Vries, Zoë Breedveld, Annelouk van Mierlo, Marnix Tijhuis, Laura Marchal-Crespo Abstract: Virtual Reality (VR) has been gaining popularity in neurorehabilitation for its potential benefits of increasing patients’ motivation and engagement [1]. An important yet relatively unexplored aspect of VR interfaces is how the patient is represented in the virtual world, e.g., with an avatar. A higher level of embodiment over avatars has been shown to enhance motor performance during upper limb training [2], [3] and has the potential to be employed to enhance neurorehabilitation [4]. However, the relationship between avatar embodiment and gait performance remains unexplored. In this work, we present the results of an experiment conducted with 12 healthy young participants to evaluate the effect of different virtual representations of the participants’ lower limbs on foot placement accuracy while stepping over a trail of 16 virtual targets. We compared three levels of virtual representation: i) a full-body avatar, ii) only feet, and iii) no representation. Full-body tracking was computed using standard VR trackers to synchronize the avatar with the user’s motions. Foot placement accuracy was measured as the distance between the foot’s center of mass and the center of the selected virtual target. Additionally, the level of embodiment of each virtual representation was evaluated through a questionnaire. We found that foot placement accuracy increased with some form of virtual representation, either full-body or foot, compared to having no virtual representation. However, the foot and full-body representations did not show significant differences in accuracy. Importantly, we found a negative correlation between the level of embodiment of the foot representation and the distance of placed foot to the target(r_Pearson=-.65,p=.02). However, no such correlation was found for the full-body representation. Our results indicate the importance of embodying a virtual representation of the foot when performing a task that requires accurate foot placement. However, showing a full-body avatar does not seem to enhance accuracy further. Furthermore, our results suggest that motor performance in this stepping task might be modulated by the level of embodiment of the virtual feet. This work motivates future research on the effect of the embodiment of virtual representations on motor control to be exploited for VR gait rehabilitation. 10:00 0 mins Age-specific topology minimization in a one-dimensional model describing carotid haemodynamics Irene Suriani, Massimo Mischi, R. Arthur Bouwman, Kevin D. Lau Abstract: Patient-specific one-dimensional (1D) haemodynamic models can support personalized clinical decisions through an improved interpretation of carotid ultrasound (cUS) velocity and diameter waveforms [1]. However, personalized 1D models of the arterial vasculature require the estimation of a large number of parameters from in-vivo measurements (e.g., vessel lengths, diameters, compliances) [2], some often not readily available in most clinical settings. In such cases, we are confronted with a problem of complexity optimization: what is the minimum required 1D topology (i.e., the number of 1D arterial branches included in the model network) such that the least possible number of parameters are to be assumed, while still being able to accurately capture characteristic cUS waveform features? In this work, pulsatile haemodynamics were simulated by solving the 1D blood flow equations with an elastic tube law using the NEKTAR++ Pulse Wave Solver. Using a systematic method for 1D model reduction, starting from a clinically validated 55-branch baseline topology [3], we have shown that the minimum topology required to accurately simulate cUS waveforms varies in virtual subjects of different ages [4]. Using this approach, we have selected age-specific reduced models that retain carotid waveform features with NRMSE <4% for velocity and <0.15% for diameter, while reducing the initial parameter set by more than half. Finally, we have used wave power analysis to elucidate the location of origin of reflected waves responsible for cUS waveform features, and investigated the effect of model reduction on these. This work sets the basis for the generation of patient-specific 1D models for the study of carotid haemodynamics in subjects of different ages. [1] I. Suriani, et al., “Carotid Doppler ultrasound for non-invasive haemodynamic monitoring: a narrative review,” Physiological Measurement, Accepted Manuscript, 2020. [2] J. Alastruey, et al., “On the impact of modelling assumptions in multi-scale, subject-specific models of aortic haemodynamics,” J. R. Soc. Interface, vol. 13, no. 119, 2016. [3] J. Alastruey, et al., “Arterial pulse wave haemodynamics,” in 11th International Conference on Pressure Surges, 2012, pp. 401–443. [4] I. Suriani, et al., “Development, validation and uses of an age-specific virtual population for the study of carotid haemodynamics,” Manuscript in Preparation. 10:00 0 mins Classification of movement disorders using freehand single camera video recordings of gait tests Wei Tang, Peter van Ooijen, Deborah Sival, Natasha Maurits Abstract: Early Onset Ataxia (EOA) and Developmental Coordination Disorder (DCD) exhibit significant phenotypic similarities that may make differentiation challenging. Gait in ataxia can be clinically assessed semi-quantitatively using the Scale for the Assessment and Rating of Ataxia (SARA). However, such clinical evaluation is subjective and requires significant clinical competence. Wearable inertial sensors might provide an objective and feasible alternative. This approach, however, needs a special setup before usage which takes time and is not very patient-friendly. Because of the widespread use of video cameras in outpatient clinics and recent advances in pose estimation, we proposed a 2D skeleton-based machine learning technique for the classification of EOA, DCD, and healthy controls from freehand single camera recordings. Eighty-five children (31 EOA, 20 DCD, and 34 controls) were asked to walk in a straight line at their own pace, towards and away from a single 2D camera. Recordings were taken at 1280 × 720 resolution and 30 frames per second. Seventeen skeleton keypoints were extracted using the AlphaPose model pretrained on the MSCOCO dataset. The PoseFlow framework was then utilized to match the skeleton to the same subject in a recording. We employed a normalization approach using the average Euclidean distance from left shoulder to right hip and from right shoulder to left hip (ASH) to derive distance-based features. To improve the classification accuracy, we employed the overlapping sliding window (OSW) method to decompose the skeleton time series into windows of 45 frames with 30 frames overlap. We used 20 combined features as obtained from principal component analysis on statistical characteristics (mean, std, cov, skewness…) of ankle, knee, hip, elbow, shoulder and wrist distances. Classification results were validated by the XGBoost model and leave-one-patient-out cross-validation. The XGBoost classifier obtained a 67% f1 score on the original data before ASH normalization and OSW data augmentation. After ASH normalization, the f1 score improved to 70%. After ASH normalization and OSW data augmentation, the classifier finally achieved an f1-score of 78% for our dataset, with 63% for EOA, 76% for DCD, and 88% for healthy control children at group-level. Our approach demonstrates a promising application for the classification of movement disorders in a practical and objective way. Additionally, the ASH normalization method and the OSW data augmentation improved the classification result. We will incorporate other tasks, assessing upper body movement, in the future to develop further improved algorithms for clinical diagnostic support. 10:00 0 mins Enhancing the shear-wave axial-velocity estimation using a convolutional neural network Xufei Chen, Nishith Chennakeshava, Rogier Wildeboer, Ruud Van Sloun, Massimo Mischi Abstract: The Signal-to-Noise Ratio (SNR) of the particle motion can directly impact the accuracy of the downstream shear wave (SW) elastography estimates. This SNR decreases when the push pulse energy is lower. However, minimizing the push energy is of clinical relevance, as lower mechanical and thermal indices ensure clinical safety. Here, we propose a 3-D multi-resolution convolutional neural network (MRCNN) to perform improved particle velocity Vz estimation, to minimize the push energy while preserving high SNR. We present a novel approach to generate training data from real acquisitions, providing high SNR targets, one-to-one paired to inputs that are corrupted with real-world noise and disturbances. 10:00 0 mins Advanced vital remote monitoring of individuals at risk of developing diabetic cardiomyotathy Yaowen Zhang, Peter H. Veltink, Dirk W. Donker, Ying Wang Abstract: Diabetes mellitus causes significant cardiovascular complications with incidences that reach pandemic proportions worldwide. Among the diabetic sequelae, diabetic cardiomyopathy has important clinical implications as it may progress towards symptomatic heart failure and irreversible cardiac remodeling. The late manifestations of diabetes strongly limit prognosis, while putting an enormous burden on health care expenses [1]. Despite the increasing awareness of disease prevention and timely medical treatment, there is still no effective way to detect diabetic cardiomyopathy at its earliest possible, potentially reversible stages. Therefore, it is urgently needed to develop advanced vital monitoring tools that allow to gain more insights into individual patient’s pathologically changed vital parameters. This could help to identify early signs of diabetic cardiovascular disease in people at a high risk of developing diabetic cardiomyopathy. To our best knowledge, such dedicated monitoring systems that could be applied for daily life usage have not been developed. From a pathophysiological perspective, early manifestations of diabetic cardiomyopathy are characterized by the imbalance of autonomic nervous system which is defined as hyperactive sympathetic activities and hypoactive parasympathetic activities as is reflected by heart rate variability (HRV) changes [2]. From this, HRV, especially the detailed analysis of its daily life associated dynamicity, can likely serve as an important candidate marker to indicate early signs of diabetic cardiomyopathy. Therefore, we aim to develop a remote monitoring system for the early detection of diabetic cardiomyopathy that can be easily applied in patients’ daily life. Our monitoring model with the inputs of daily physical activity, such as, walking, running, and cycling, should ultimately predict cardiac responsiveness to the individual's activity (notably HRV) and detect 'warning patterns' of patients’ actual cardiac responses. The wearable sensors used to track body movement signals will be inertial measurement units (IMUs), which contain three-axis accelerometers and three-axis gyroscopes. The cardiac activity prediction model will be built as a combination of both physiological models and data driven models. Decision support algorithms to analyze disease risks will be generated as based on the data derived from pre-diabetic and diabetic individuals in order to derive early 'warning patterns' of cardiac responses. References [1] Jia, G., Hill, M. A., & Sowers, J. R. (2018). Diabetic Cardiomyopathy: An Update of Mechanisms Contributing to This Clinical Entity. Circulation research, 122(4), 624–638. https://doi.org/10.1161/CIRCRESAHA.117.311586 [2] P. Van Kessel, D. De Boer, M. Hendriks, and A. M. Plass, “Measuring patient outcomes in chronic heart failure: Psychometric properties of the Care-Related Quality of Life survey for Chronic Heart Failure (CaReQoL CHF),” BMC Health Serv. Res., vol. 17, no. 1, pp. 1–7, Aug. 2017, doi: 10.1186/S12913-017-2452-4/FIGURES/1. 10:00 0 mins Design, optimization and sensorization of a 3D printed surgical manipulator Izadyar Tamadon, Gabrielle Tuijthof, Arianna Menciassi Abstract: Designing efficient flexible manipulators in Minimally Invasive Surgery (MIS) demands high maneuverability, accuracy, and safety. Current handheld manipulators in the state of art are limited in flexibility and tip’s force/position information. In this work, we propose a 3D printed handheld steerable manipulator with high torsion and axial stiffness as well as low bending stiffness by combining four helicoids and a continuum backbone. Simulation results in SOLIDWORKS software show that simplified static modeling can be used to analyze the distributed stress in the helical design. Design optimization has been subsequently conducted to investigate the structural parameters for further performance improvement. In order to steer this flexible manipulator, a control handle was designed by combining commercial compression spring and 3D printed backbones. Moreover, to estimate the tip position and external forces on the tip, inductance sensors in the form of commercial tension springs were combined inside the cable actuation design. On-bench experiments have been conducted to analyze and validate the bending and loading performances with respect to kinematic models. The manipulator was equipped with surgical grasper to perform pick and place experiments in a confined space, indicating great potential for the applications in MIS and endoscopic procedures. The experimental results show that the proposed handheld manipulator achieves acceptable distal end positioning accuracy with an average error of 7%. Meanwhile, the proposed sensing can estimate the static payload on the tip with an average accuracy of 5%. We showed the importance of stress-strain simulations while prototyping steerable surgical instruments by brittle 3D printer materials and the efficient sensory data to a successful navigation and load estimations on the distal tip. 10:00 0 mins Development of non-invasive method to measure the mechanical properties of skin using ultrasound techniques Zülal Kizilaslan, Marcel Rutten, Frans van de Vosse, Richard Lopata Abstract: Measuring the biomechanical change of skin can be used to detect skin diseases and monitor the development of skin problems such as pressure ulcers [1]. The objective of this study is to develop an experimental setup to estimate the mechanical properties of skin based on quasi-static ultrasound (US) elastography, a technique used to estimate biomechanical quantities and properties of tissue (such as strain and modulus), non-invasively [2]. For an in vitro feasibility test, phantoms were created using a 15 weight percent (wt%) polyvinylalcohol (PVA) solution subjected to 1 to 3 freeze-thaw cycles. For acoustic scattering, 3 weight percent silicon carbide was added to the PVA mixture. The skin mimicking phantoms were deformed while performing ultrasound imaging. To create a heterogeneous deformation field and measure the load exerted by the ultrasound probe, a water-filled, small diameter latex balloon (Cattex, Casalvieri, Italy) was used. This balloon was attached to a pressure sensor and positioned between a linear array probe (Esaote Europe, 10MHz) and the upper surface of the phantom. During compression, radio frequency data were recorded and exported to a PC. Data were processed for image reconstruction, and strain imaging was performed offline. Local strains were determined for each frame using a 2-D block matching technique, both implemented in MATLAB. The pressure inside the balloon was measured using a pressure sensor (P10EZ, BD, Vianen, the Netherlands). A Finite Element (FE) Model was developed to verify the US results. The 3D FE model consists of the PVA phantom, the water-filled balloon, and the probe as indenter, all implemented in ABAQUS (Dassault Systèmes, France). Displacement and strain fields were estimated throughout the compression experiments. The FEA shows similar deformation patterns as measured with ultrasound. However, the water pressure is currently still overestimated which might be caused by the current contact definition. In future work, the FEA model will be improved further, and the combination of measurement and model will be used to characterize the local mechanical properties of the phantom and eventually those of skin tissue. 10:00 0 mins Prediction of thrombosis at a stenotic region due to neointimal hyperplasia Mohammad Rezaeimoghaddam, Bregje in 't Veld, Lotte Piek, Frans van de Vosse Abstract: Autologous veins are widely used in bypass surgeries because of their availability, ease of use, and lower infection risk compared to synthetic grafts. However, the poor long-term outcomes of venous grafts (VGs) are mainly due to the luminal narrowing that results from neointimal hyperplasia (NIH) and subsequent superimposed thrombosis. Hence, the main challenge is detection of progressive vascular wall thickening due to NIH and avoid sudden thrombus formation. In spite of decades of experience in animal experiments, vascular tissue engineering design, and experimental trial-and-error, novel approaches are urgently required. Computational models are powerful tools that can provide great support and information in medical research. In this study two major processes responsible for VG failure, are analyzed using a computational model of thrombus formation in relation with NIH. In this study, a novel continuum model of platelet activation, aggregation, and essential regulators of the coagulation cascade is used to describe transport and reaction of biochemical agonists and predict thrombosis. The current model includes the features of our previous deposited bounded platelet model including shear induced platelet activation and embolization [1]. This model also contains von Willebrand factor (vWF) conformational change which stretched vWF can be bound to platelets integrins and collagen. The NIH model is based on the conservation of mass between VSMCs and the migration of cells between media and intima. The VSMCs phenotypical change from quiescent cells, migration from the media, proliferation and apoptosis are solved with a system of ordinary differential equations. A moving boundary technique is used to change computational nodes and alter the lumen diameter corresponds to NIH at each time step. The computational domain consists of a 2D channel with a semi-ellipse which represents stenosis at the bottom wall. This computational model will enable identifying key factors associated with the VG failure, resulting in new insights that are critical for guidelines to ensure patency of grafts. The results also accurately describe the interactions between the regulatory network of the coagulation cascade, agonist concentrations and dynamics of platelet deposition and fibrin due to hemodynamics. Ref: [1] M Rezaeimoghaddam and FN van de Vosse. J. Biomech. 2022. 10:00 0 mins Combining haptics and immersive virtual reality in gait robot-assisted rehabilitation Stefano Dalla Gasperina, Alex van den Berg, David Abbink, Laura Marchal-Crespo Abstract: In the past decades, gait robotic devices, e.g., treadmill-based exoskeletons, have been employed to deliver more cost-effective and intensive rehabilitation training after stroke. Robots can supplement therapists during training, relieving their workload and allowing high-intensity repetitive movements. Virtual reality (VR) and serious games can be concurrently employed to enhance the patients’ motivation [1]. However, in current approaches, these technologies are merely superimposed and the VR is usually presented on a flat-screen [2]. Within this project, we aim to combine gait robotic devices with more immersive VR techniques by providing concurrent haptic and visual feedback to enhance motivation, engagement, effort, and movement variability, all aspects suggested to improve neurorehabilitation outcomes [3]. Our idea is to employ a customized gait exoskeleton to provide the haptic rendering of the interaction with the virtual environment – i.e., by simulating the interaction forces with tangible virtual objects according to their dynamic model. Simultaneously, we aim to show congruent visual information through head-mounted displays (HMDs). We will first enable our treadmill-based exoskeleton to serve as a haptic display, i.e., being able to generate external perturbations to the natural gait that can be mapped into tangible virtual objects or environments, such as walking into the water. To enhance the level of presence in the virtual environment, we will use HMDs to show intuitive realistic virtual environments that can be coherently synchronized with the interaction forces generated by the exoskeleton. Additionally, a virtual avatar of the patient will be synchronized with the gait pattern. Finally, we will evaluate the effectiveness and acceptance of the proposed approach by comparing motor performance, user experience, and motivation under four different conditions – i.e., combining (or not) haptics and IVR. Our research has great potential to impact the field of robot-assisted neurorehabilitation. We expect our results to confirm that combining haptics and immersive VR can enhance user motivation and participation during training. Further, our approach can broader the possibilities of using robotic devices in the clinic, as they could be used in patients with a large range of disability levels, and facilitate the transfer of the learned gait patterns to activities of daily living. [1] N. Wenk, K. Buetler, J. Penalver-Andres, R. Müri, and L. Marchal-Crespo, “Naturalistic visualization of reaching movements using head-mounted displays improves movement quality and proves high usability compared to conventional computer screen.” 2022. doi: 10.21203/rs.3.rs-1749644/v1. [2] N. Hamzeheinejad, D. Roth, S. Monty, J. Breuer, A. Rodenberg, and M. E. Latoschik, “The Impact of Implicit and Explicit Feedback on Performance and Experience during VR-Supported Motor Rehabilitation,” in 2021 IEEE Virtual Reality and 3D User Interfaces (VR), Mar. 2021, pp. 382–391. doi: 10.1109/VR50410.2021.00061. [3] Ö. Özen, K. A. Buetler, and L. Marchal-Crespo, “Promoting Motor Variability During Robotic Assistance Enhances Motor Learning of Dynamic Tasks,” Front. Neurosci., vol. 14, 2021. 10:00 0 mins Cardiac resynchronisation therapy vs Left Bundle Branch Area Pacing: a model study Roel Meiburg, Jesse Rijks, Ahmed Beela, Justin Luermans, Luuk Heckman, Tammo Delhaas, Frits Prinzen, Kevin Vernooy, Joost Lumens Abstract: Cardiac resynchronisation therapy (CRT) has shown to reduce mortality in patients suffering from heart failure, but still leads to non-physiological electrical and mechanical activation of the heart. Recently, alternative pacing strategies such as left bundle branch pacing (LBBP) and LV septal pacing (LVSP) are being explored, as they are thought to be more physiological. Although much work has been done to explore the electrophysiological response to LBBP/LVSP, their effect on left and right ventricular mechanical and haemodynamic function is underappreciated. This is partially due to the fact that cardiac function is also highly dependent on the mechanical viability of the underlying tissue, which is difficult to measure clinically and often already compromised in heart failure patients. Computational modelling provides an opportunity to help understand and directly compare the effects of LBBP/LVSP and CRT without any confounding variables. Furthermore, the effect of each pacing strategy can be simulated for different heart failure phenotypes. To simulate the placement of a pacing device, electrical activation maps were generated on a realistic 3D geometry of the LV and RV using a graph-based formulation of the Eikonal model. The geometry was segmented and activation times for each segment were averaged, which were then used to inform the CircAdapt model of human cardiovascular mechanics and haemodynamics to calculate the effects of the different pacing strategies on cardiac pump function. This approach is computationally cheap compared to full 3D (Finite Element) simulations, whilst still providing results on a spatial resolution similar to data currently available in the clinical setting. Initial results show that selective LBBP mostly restores LV function at the cost of RV function loss, similar to a right bundle branch block pattern. Non-selective LBBP reduces RV load compared to sLBBP, but leads to earlier septal activation and increased intra-ventricular dyssynchrony. LVSP further reduces RV load and increases interventricular dyssynchrony, but still performs similarly to CRT. These findings hold when simulating different levels of mechanical viability and are consistent with clinical findings. Finally, this framework is easily expanded to simulate other cardiac dysfunctions, such as myocardial infarction. 10:00 0 mins Estimation of joint moments using IMUS to aid clinical decision making during ACL rehabilitation: A review Sanchana Krishnakumar, Bert-Jan van Beijnum, Jaap Buurke, Peter Veltink, Chris Baten Abstract: Rehabilitation after an anterior cruciate ligament (ACL) injury is divided into multiple phases and progress between phases is based on the functional assessment of patients. These assessments are currently subjective and are done by visual monitoring of activities such as running, hopping, jump landing, and cutting maneuvers by physiotherapists. Estimation of objective measures like knee joint moments and ground reaction forces (GRF) during assessment can help in gaining new insights on knee loading and open new avenues for rehabilitation. Accurate estimation of kinetics is a complex task and requires expensive motion capture systems along with force platforms. On the other hand, several algorithms have been proposed in the literature to estimate kinetics by just using kinematics measured with inertial sensors (IMUs). However, the knowledge about their clinical applicability is limited. This study aims to compare available algorithms for the prediction of GRF and/or estimation of joint moments only using IMUs and evaluate their feasibility for adoption in ACL rehabilitation. A literature search was conducted using Scopus (Elsevier), PubMed and SportDiscus with keywords including ‘GRF’, ‘Joint kinetics’, ‘IMU’ and commonly accepted variations of these terms. The identified studies were classified based on the parameters estimated (joint moments/GRF) and the principles used such as machine learning (ML), musculoskeletal modelling, hybrid, direct modelling, or statistical approach. The comparison of the algorithms was done based on the accuracy achieved, assumptions used, tasks validated and their applicability for ACL patients. Most of the studies evaluated have estimated only vertical GRF with good accuracy and reported lateral GRFs as less reliable. ML-based approaches have proved to be more versatile but have the disadvantage of sensitivity to input parameters and require large sets of training data. Tasks such as walking, and landing also involve double support phases where further transfer functions are required to distribute forces between the legs. The applicability of assumptions made for distribution is unclear for ACL patients. The assessed algorithms have also not yet been widely validated for tasks such as jump landing and hoping. A combination of two methods such as biomechanical modelling and ML or musculoskeletal modelling may be used to further increase accuracy and make them versatile to estimate joint moments for a large range of movements. Further validation and tuning of these algorithms are thus necessary before being implemented for ACL monitoring and phase decision. 10:00 0 mins Towards personalized monitoring of carotid plaque growth and rupture-risk using ultrasound-based flow imaging Janna Ruisch, Suzanne Holewijn, Michel Reijnen, Chris de Korte, Anne Saris Abstract: Rupture of a carotid plaque may lead to stroke. Patient selection for surgical plaque removal, carotid endarterectomy, is far from perfect; the number-needed-to-treat in patients with severe stenosis is six to prevent one stroke within 5-years. Ultrasound is, together with CT-scanning, the technique of choice for treatment planning, in which simple geometrical and blood velocity measures are used. However, conventional ultrasound-based velocity measurements assess the along-the-beam velocity component, resulting in large variability. Together with the low frame rates, this restricts measurement of complex or high blood velocity. Especially these complex flow patterns, and the resulting wall shear stresses (WSS) acting on the vessel wall, are crucial for plaque initiation, growth and rupture. The advent of ultrafast, i.e. high-frame-rate, ultrasound makes continuous tracking of flow in all directions feasible. High-frame-rate blood flow imaging (BFI) in the carotid artery is possible by using cross-correlation-based speckle tracking. In this project, we aim to assess if and which blood flow-related parameters, such as vorticity, vector complexity and WSS, are associated with plaque growth and vulnerability using non-invasive, ultrasound-based BFI. In two ongoing clinical studies, the relation between flow patterns and plaque growth and vulnerability will be explored. BFI is performed in asymptomatic patients (N=11/85) during 2-year follow-up to evaluate the association between flow patterns and plaque growth. Furthermore, BFI is acquired in patients (N=5/70) scheduled for carotid endarterectomy. Histology staining of the excised plaque will differentiate vulnerable from stable plaques. Vector complexity and vorticity were derived from velocity fields at pre-, max- and post-stenotic regions. Vector complexity (-) defines the spread of velocity vector directions in a region of interest (ROI). Vorticity (rad/s) is the local rotation of blood within an ROI. Preliminary results show the ability to discriminate different regions in stenosed carotid arteries based on these blood flow-related parameters: increased vector complexity was observed at post-stenotic regions compared to pre- and max-stenosis, especially during the systolic phase. By quantifying different blood-flow related parameters from ultrasound-based BFI, we aim to demonstrate the potential of this novel technique to predict plaque growth and vulnerability. In future, this might enable more sophisticated, personalized treatment decision-making. 10:00 0 mins A human disease model for atherosclerosis – the first steps towards understanding atherosclerotic plaque rupture Tamar Wissing, Sheila Serra, Anthal Smits, Carlijn Bouten, Frank Gijsen, Kim van der Heiden Abstract: Stroke, one of the leading causes of death worldwide, is caused by rupture of the fibrous cap overlying the atherosclerotic plaque in the carotid artery. Cap rupture is difficult to predict due to the heterogenous composition of the plaque, unknown material properties, and the stochastic nature of the event. We aim to create a tissue engineered human disease model for atherosclerosis with a variable but controllable collagen composition, suitable for mechanical testing, to scrutinize the reciprocal relationships between biological composition and mechanical properties to unravel rupture mechanics. Human Vena Saphena Cells (HVSCs) were cultured in 1 x 1.5 cm-sized fibrin-based constrained gels for 21 days according to previously established [1] dynamic culture protocols (i.e. static, intermittent or continuous loading) to vary collagen composition (e.g. amount, type and organization). After 7 days of static culture, a soft 2 mm ∅ fibrin inclusion was introduced in the centre of each tissue to mimic the soft lipid core, simulating the heterogeneity of a plaque. The samples were statically cultured for another week, whereafter they were exposed to an intermittent or continuous straining protocol up till 21 days using the Flexcell FX-40001 (Flexcell Int, McKeesport, PA). Statically cultured samples were included as controls. Samples were exposed to uniaxial tensile tests or analyzed via imaging and immunohistochemistry (IHC) at day 21 to determine mechanical properties and collagen composition (e.g. organization, quantity & type), respectively. Results demonstrate reproducible collagenous tissues, that vary in collagen composition due to the presence of a successfully integrated soft inclusion and the culture protocol applied. All statically cultured samples exhibited an isotropic collagen organization, irrespective of the locus analysed. In contrast, both the samples that were exposed to intermittent as well as the continuous loading demonstrated an isotropic to anisotropic collagen configuration when moving from the top of the construct to the shoulder and mid region. The model mimics the bulk mechanical properties of human caps and can now be deployed to further assess tissue mechanical properties and failure of fibrous caps to better understand fibrous cap rupture and to identify new biomarkers for identification of plaques at risk of rupture. 10:00 0 mins Procedure specific radiation analysis during vascular and trauma surgery Teddy Vijfvinkel, Caroline Janssen, Karlijn Scheepens, Vincent Verhoeven, Benno Hendriks, John van den Dobbelsteen, Maarten van der Elst Abstract: Introduction In current vascular and trauma surgery, imaging techniques using X-rays are an indispensable part of the workflow. At the same time, radiation creates a risk for both patient and medical staff. The ALARA principles provide measures to keep radiation exposure as low as reasonably achievable. However, procedure-specific radiation load is unknown and targeted feedback is non-existent. Therefore, this study aimed to determine the radiation exposure of medical staff in order to understand adherence to guidelines and target areas for improvement. Methods A single-site prospective observational cohort study was conducted. Data on scattered radiation exposure was collected using Philips DoseAware dosimeters during vascular and trauma surgery requiring X-ray. Surgeons were equipped with dosimeters on the forehead and on both wrists; all other staff wore dosimeters on top of the lead apron at chest height. The total radiation dose used during surgery was retrieved from the C-arm. A ratio between used and received radiation was then calculated to compare procedures. Results During this study, 84 procedures were included consisting of 71 trauma surgeries and 13 vascular surgeries. The dosimeter on the left wrist of the primary surgeon received the highest radiation dose compared to the other dosimeter locations: 23,58 µSv (SD ± 31,33) vs 3,81 µSv (SD ± 7,65), p<0,001. The average dose received by the primary surgeon was 12,12 µSv (SD ± 24,97) per procedure, whilst the secondary surgeon received on average 4,72 µSv (SD ± 7,11), p<0,001. The anaesthesia assistant was exposed to the smallest radiation dose with an average of 0,58 µSv (SD ± 1,43) per procedure. All staff members received significantly more radiation during vascular surgery compared to exposure during trauma surgery, although this difference reduced when correcting for the total radiation dose used. Discussion This study gives a unique insight in the distribution of scattered radiation during vascular and trauma surgery. It clearly shows that the primary surgeon, and specifically the left arm, receives the most scattered radiation throughout the procedures, albeit acceptable and still far from the annual threshold of 20 mSv. Furthermore, this study confirms a significantly higher radiation dose is used and received during vascular surgery in comparison to trauma surgery. With these insights, procedure-related feedback systems can be developed to take specific precautions and enhance radiation awareness and hygiene. 10:00 0 mins Towards mechanical characterization of both AAA wall and intraluminal thrombus using 3D+T ultrasound Arjet Nievergeld, Judith Fonken, Esther Maas, Marc van Sambeek, Richard Lopata Abstract: Introduction An abdominal aortic aneurysm (AAA) is a localized dilatation of the aorta, which in case of rupture has a mortality rate of 80%. Current clinical guidelines of intervention are based on AAA diameter, which has been proven to be an inadequate criterion. Biomechanical models can improve rupture risk prediction in a more patient-specific way, using e.g. CT or ultrasound (US) imaging [1, 2]. US is safer compared to CT and adds temporal information for mechanical characterization of the AAA. It is hypothesized that the intraluminal thrombus (ILT) lowers the wall stress and therefore should be included in rupture risk assessment [3]. The objective of this study is to show feasibility to use the acquired time-varying geometry of the lumen and vessel wall for mechanical characterization of both ILT and AAA wall, for future patient-specific modelling of the AAA, including ILT. Materials and Methods An in-house developed semi-automatic segmentation method was used to segment the lumen, ILT, and vessel wall out of 3D+t US images. Despite the low contrast, the lumen diameter-time curve showed a clear blood pressure induced pulsatility in 22 out of 26 patients. Those diameter-time curves were used to calculate the ILT systolic-diastolic volume change, and both lumen and vessel distensibility [4]. Results The ILT volume change (median -0.57%) was within the IQR range of Kontopodis et al. [5]. The median distensibility of the lumen (5.9 MPa-1) and vessel wall (2.8 MPa-1) are lower than the ones found in Molacek et al. (lumen 12 - 30 MPa-1, vessel 5 – 13 MPa-1)[4]. Discussion and Conclusions The compressibility was in agreement with literature showing that ILT can be considered as incompressible. The under-estimation in distensibility might result from the low frame rate of (3D+t) US (3.2-7.3 Hz), since this might cause an under-estimation in systolic diameter and over-estimation of the diastolic diameter (Figure 1). Despite this limitation, a higher distensibility was found for the lumen, compared to the vessel. This observation is in agreement with Molacek et al. [5]. Future research will focus on incorporating these mechanical properties in personalized biomechanical models to improve AAA rupture risk assessment. References 1. Van Disseldorp et al, Eur J Vasc Endovasc Surg, 59: 81-91, 2020 2. Kok et al, J Vasc Surg 61: 1175-1185, 2015 3. Domonkos et al. , Int. Angiol.38: 39-45, 2019 4. Molacek et al. Ann Vasc Surg 2011;25:1036-1042 5. Kontopodis et al. Theor Biol Med Model. 2013;10(1):1-9 Acknowledgements: This work was supported by e/MTIC MEDICAID project.

end %-->