Keywords: Artificial organs
Abstract: Compared to other heart valves, the aortic valve is particularly susceptible to disease with nearly 1% of the US population having either aortic stenosis or aortic regurgitation. Since 1960, artificial heart valves have been used to replace diseased native valves and have saved millions of lives. Mechanical valve designs include caged-ball, tilting-disc, and bileaflet valves; surgical bioprosthetic designs include porcine and pericardial stented and non-stented valves. Trans- catheter aortic heart valves, a more recent innovation in artificial heart valve technology, bring the benefit of heart valve replacement to higher surgical risk patients. Unfortunately, despite over five decades of use and innovation, these prosthetic heart valves are less than ideal and lead to many complications including thrombosis, hemolysis, paravalvular leakage, conduction abnormalities, and prosthesis-patient mismatch. Many of these complications/problems are directly related to the fluid mechanics associated with the various valve designs. To shed light on the mechanisms behind these problems, state-of-the-art experimental and computational fluid dynamics (CFD) approaches have been employed. These approaches involve the use of left heart simulators, particle image velocimetry, echocardiography, laser Doppler velocimetry, flow visualization, multi-scale CFD modeling, and fluid-structure interaction simulations. Detailed flow fields obtained have facilitated in-depth analyses of shear stress, vorticity, and residence times, markers for hemolysis, thrombosis, and other complications. These studies have yielded a better understanding of how prosthetic heart valves in current clinical use may create non- physiologic and adverse flow fields. All in all, potential clinical failure mechanisms could be averted via a holistic approach to flow studies.

Keywords: Biological and medical system modelling, Biomechanics, Cardiovascular system
Abstract: The purpose of this work was to know the characteristics of tricuspid bioprosthetic mitral valves in terms of velocity fields and valve aperture areas for some instants concerning the diastole. A left heart simulator was used as an in vitro platform for the bioprostheses. A particle image velocimetry (PIV) system was applied for characterizing instantaneous 2D-velocity fields within the left ventricular model (LVM), and also for detection of mitral bioprostheses aperture areas. The ventricular states were established for 60, 70 and 90 bpm (beats per minute) and a blood analog fluid based on glycerol was employed at 36.5±0.5ºC. Two mitral bioprostheses (nominal diameter of 27 and 31 mm) and an aortic bileaflet mechanical valve (27 mm diameter) were used. Instantaneous two-dimensional velocity fields at the upper part of the LVM, for two sections (in the middle of the LVM, and also in a section at 7 mm of distance from that) were obtained. Averaged velocity vector maps (N = 50) were calculated for some instants of interest. According to different heart rates, bioprostheses and sections within the LVM, the velocity fields were presented, as well the bioprostheses aperture areas. In conclusion, the experimental scope including different ventricular states showed how the inflow characteristics of the LVM change according to the chosen variables. This information will aid the validation of a numeric model for the LVM operation, which has been elaborated (some specific methods are presented in the Appendix).

Keywords: Biosignal analysis, processing and interpretation, Cardiovascular system, Decision support systems
Abstract: This paper develops a novel methodology for extracting simple, beat-to-beat, lumped metrics of systemic circulation performance by comparing the catheter pressure waveforms from the femoral artery and vena cava. Their diagnostic potential is compared to the similar, clinically used metric ΔMP, the drop in mean blood pressure between the aorta and vena cava, across an experimental cohort encompassing the progression of sepsis and several clinical interventions known to alter circulatory state. Both Ofe→vc, the model derived pressure attenuation between the femoral artery and vena cava, and the clinical metric, ΔMP, performed well. However, Ofe→vc reduced the optimal average time to sepsis detection from endotoxin infusion from 46.2 minutes for ΔMP to 11.6 minutes, for a slight increase in false positive rate from 1.8% to 6.2%. Thus, the potential diagnostic benefits of this novel approach are demonstrated, and a case is made for further investigation.

Keywords: Biosignal analysis, processing and interpretation, Cardiovascular system, Medical technology
Abstract: Pulse oximeters are frequently used to provide real time measurements of heart rate and blood oxygen saturation (SpO2). SpO2 is calculated by taking the ratio of the AC to DC components of the photoplethysmograph (PPG) signal measured by the pulse oximeter. For accurate estimation of SpO2, the AC component needs to be extracted from the signal through signal processing, where accurate peak detection is a crucial, difficult element. This paper investigates the use of the wavelet transform for real time signal processing to detect peaks that could be unintentionally attenuated through more conventional filtering methods. Four mother wavelets (Daubechies 3, symlets 2, coiflets 3 and reverse biorthogonal 1.5) were tested against each other to determine the wavelet with the best representation of the PPG signal in a noisy environment (SNR of 6.44). The reverse biorthogonal (rbio1.5) mother wavelet was found to better represent the PPG signal with a specificity of 0.97 and a sensitivity of 0.97. Further research into the decomposition depth of the rbio1.5 wavelet resulted in an optimal depth of 3, with the 2nd and 3rd levels being used for reconstruction of the signal. Using a wavelet length of 128 samples resulted in a time delay of 2.56 seconds. This time delay is well within clinical requirements for near real-time-signal analysis involving these devices.

Keywords: Control of medical devices, Control of physiological and clinical variables, Biological and medical system modelling
Abstract: Norm-optimal iterative learning control algorithms use plant models to predict the system behavior. In this paper, we focus on improving the performance in the norm-optimal iterative learning control of left ventricular assist devices. For this purpose, the previously used simple plant model is replaced by a piecewise linearized version of a nonlinear cardiovascular system model including the left ventricular assist device. Simulations are carried out to study the controller response to end-diastolic volume setpoint and preload changes. The results show minor improvements regarding the tracking performance and the rejection of disturbances but also an increase of computational effort compared to the previous algorithm.

Keywords: Simulation and visualization, Biomechanics, Cardiovascular system
Abstract: The main cause of death in most countries of the world is cardiovascular diseases. Heart transplant or/and the usage of a Ventricular Assist Device (VAD) is often a form of treatment for severe heart diseases. The objective of this study is to analyze the importance of volute vanes in the efficiency of a new type of VAD, the Transventricular Assist Device (TVAD). Investigations concerning the hydrodynamic performance are conducted using finite element methods aiming at the best conditions to support the circulatory system and to avoid to the maximum areas of turbulent flow. Two three-dimensional pump models were created using computer aided design system, one without and the other with volute vanes. Flow in both models was, then, analyzed by Computer Fluid Dynamic (CFD) tools. Analyses showed that, when volute contains vanes, flow is redirected toward the rotor while, in the case of volute without vanes, the fluid recirculates in the pump's interior. From now on, TVAD studies will always consider the presence of vanes in the volute.

Keywords: Biological and medical system modelling, Biosignal analysis, processing and interpretation, Cardiovascular system
Abstract: Continuous cardiac output monitors are becoming more common in clinical settings to assess cardiac performance. Pulse contour analysis is a common method employed by commercial devices to estimate patient hemodynamics from a pressure waveform and relate it to volume. The main issue with current devices, is they can perform poorly during and after a significant hemodynamic event. An existing pulse contour analysis method, under ideal experimental conditions, demonstrated the ability to track changes in stroke volume (SV) using a measure of pulse wave velocity (PWV). In this study, the existing method's ability to estimate SV was tested during vena cava occlusions (VCO), a textit{worst case}, rapid transient hemodynamic change. Additionally, the method's sensitivity to the location of the arterial pressure waveform measurement was also assessed, by comparing SV estimates from aortic and iliac pressures, to SV measured by admittance catheter in the ventricle. Results show the model accurately tracks changes in SV as a result of the occlusion, a significant improvement over current commercially available devices. Bland-Altman analysis showed no significant improvement in SV estimation when using aortic pressure compared to the iliac pressure waveform, with mean bias of -2.11ml and 0.13ml, respectively. This is a desirable result, as more distal arterial pressure measurement locations increase the clinical feasibility of the method.

Keywords: Biological and medical system modelling, Diabetes management, Medical technology
Abstract: Abstract: The aim of this study is to develop an algorithm for detection of unannounced meals and an insulin bolus calculator (BC) to work in combination with the meal detector. The input of the meal detector are the continuous glucose monitoring (CGM) data and the insulin infusion rate. During daytime, the automated meal detector and the BC control the blood glucose concentration. During nighttime, a model predictive control (MPC) algorithm regulates the basal insulin rate. The meal detector detects the occurrence of a meal, estimates the amount of carbohydrate (CHO) in the meal, and estimates the meal onset time. The BC computes a bolus dose to cover the detected meal. We test the meal detector and the BC on nine virtual type 1 diabetes (T1D) patients. The meal detection algorithm, applied on the virtual patients, has a median detection delay of 40 min, detection sensitivity of 80% and a median meal onset estimation bias of 15 min. The algorithm does not have false positive.

Keywords: Diabetes management, Artificial organs, Control systems
Abstract: In this paper, we evaluate the closed-loop performance of two switching strategies for a dual-hormone artificial pancreas (AP). The dual-hormone AP administers insulin and glucagon subcutaneously. Since insulin and glucagon have opposite effects, we want to avoid simultaneous injections of these two hormones. To handle non-simultaneous injections of insulin and glucagon, we compare model predictive control (MPC) algorithms using a hysteresis switch between insulin and glucagon controllers with a multiple-input single-output (MISO) formulation. Although the closed-loop performance of these two control strategies is similar, the hysteresis switch is preferable due to (i) its greater flexibility in control design and tuning and (ii) a more straightforward way to avoid simultaneous injections of insulin and glucagon.

Keywords: Diabetes management, Control of medical devices, Pharmacokinetics and drug delivery
Abstract: People with diabetes mellitus type 1 could benefit from fully automated systems for glucose control. However, faults in any component of the system can severely compromise the safety of the user. An increasing degree of automation also increases the risk that faults remain undiscovered for longer periods - unless automated routines for fault detection are implemented at the same time. The aim of this article is to give a categorised overview of methods for fault detection in glucose control systems. This overview targets at disclosing hidden potentials for improvement and unresolved issues. Methods for fault detection in glucose control systems have been reviewed and classified with respect to categories such as the type of method and the exploited data basis. Both journal and conference papers were taken into account. Compared to the number of studies on glucose control algorithms, only a few articles have been published on fault detection. Surprisingly few of them consider system information beyond the standard diabetes care data.

Keywords: Diabetes management, Simulation and visualization, Decision support systems
Abstract: Patients with type 1 diabetes mellitus (T1DM) typically determine their bolus insulin needs based on methods from Advanced Carbohydrate Counting (ACC). In ACC the amount of required bolus insulin is assumed to be directly proportional to the carbohydrate content of the ingested meal. It is well known that many T1DM patients have diffculties to accurately estimate meal carbohydrates. Though these estimates are used for calculating the meal bolus amount, there is a scarcity of data on how estimation errors affect glycemic control. The current paper analyzes the effect of carb counting errors on glycemic outcomes during basal-bolus-therapy in a simulation study using so-called Deviation Analyses. Furthermore, the effect of inaccurate estimates on the settings of the insulin therapy is studied by means of the previously published Adaptive Bolus Calculator (ABC) algorithm. It is found that whereas random carb counting inaccuracies indeed do lead to an inferior glycemic control, systematic biases in the estimates are expected to hardly affect the results since these are usually implicitly accounted for in the therapy settings.

Keywords: Control systems, Diabetes management, Simulation and visualization
Abstract: In this paper, we evaluate the closed-loop performance of a control algorithm for the treatment of type 1 diabetes (T1D) identified from prior continuous glucose monitor (CGM) data. The control algorithm is based on nonlinear model predictive control (NMPC). At each iteration, we solve an optimal control problem (OCP) using a sequential quadratic programming algorithm with multiple shooting and sensitivity computation. The control algorithm uses a physiological model of T1D to predict future blood glucose (BG) concentrations. The T1D physiological model takes into account the dynamics between subcutaneously administered insulin and blood glucose, the contribution of meal absorption and the lag and noise of CGM measurements. The model parameters have been identified using prior data. Numerical simulations on 10 patients show that the NMPC algorithm is safe and is able to optimize the insulin delivery in patients with T1D.

Keywords: Critical care, Control of physiological and clinical variables, Metabolic system
Abstract: Glycaemic control in intensive care unit has been associated with improved outcomes. Metabolic variability is one of the main factors making glycaemic control hard to achieve safely. STAR (Stochastic Targeted) is a model-based glycaemic control protocol using a stochastic model to predict likely distributions of future insulin sensitivity based on current patient-specific insulin sensitivity, enabling unique risk-based dosing. This study aims to improve insulin sensitivity forecasting by presenting a new 3D stochastic model, using current and previous insulin sensitivity levels. The predictive power and the percentage difference in the 5th-95th percentile prediction width are compared between the two models. Results show the new model accurately predicts insulin sensitivity variability, while having a median 21.7% reduction of the prediction range for more than 73% of the data, which will safely enable tighter control. The new model also shows trends in insulin sensitivity variability. For previous stable or low insulin sensitivity changes, future insulin sensitivity tends to remain more stable (tighter prediction ranges), whereas for higher previous variation of insulin sensitivity, higher potential future variation of insulin sensitivity is more likely (wider prediction ranges). These results offer the opportunity to better assess and predict future evolution of insulin sensitivity, enabling more optimal risk-based dosing approach, potentially resulting in tighter and safer glycaemic control using the STAR framework.

Keywords: Biomedical imaging systems and image processing, Respiration and ventilation
Abstract: 3D-EIT imaging is already a reality at the bedside. Using a single and ergonomic electrode belt with special disposition of electrodes, it is possible to reconstruct multiple cross-sectional planes of the thorax along the craneo-caudal axis. One of the key concepts of this technology is the use of a statistical 3D-anatomical atlas of the thorax, based on a bank of CT images of patients with different anthropometric characteristics. The resolution within the z-axis (craneo-caudal) depends on the disposition of electrodes, but it is enough to provide independent information from different lobes of the lung. This special characteristic is important during the monitoring of localized diseases, like in patients with localized emphysema or localized pneumonia. Fast real-time image reconstruction is possible with current processors, providing the characteristic high-temporal resolution of EIT. The study of cardiac movements and cardiac output is a promising area of application for this new technology. The distribution of inhalatory drugs, like surfactants, is another promising application that will be discussed during this presentation.

Keywords: Biological and medical system modelling, Respiration and ventilation
Abstract: Models of respiratory mechanics can be used to titrate patient-specific mechanical ventilation (MV) settings in critical care, but often perform poorly in the presence of patient breathing effort. Respiratory mechanics are conventionally calculated using only inspiratory data. Muscle activity is normally assumed relatively minimal or absent during passive expiration regardless of the presence of inspiratory spontaneous breathing (SB) efforts. Hence, this study assesses whether expiratory lung elastance can be used to estimate inspiratory lung elastance for spontaneously breathing, reverse triggered patients. Clinical data from recruitment manoeuvres in fully sedated patients were used to determine a relationship between inspiratory and expiratory modeled lung elastance. The validity of this relationship was assessed using data recorded pre- and post- sedation from different patients.

A strong, linear relationship was found between inspiratory and expiratory elastance in fully sedated patients, with gradient 1.04~[95%~CI:~1.03-1.07] and intercept 1.66~[1.06-2.08] with R^2 = 0.94. After adjustment according to the linear relationship, expiratory elastance produced stable estimations post sedation, with similar median and variance as inspiratory elastance. However, variation in estimates pre-sedation, although significantly improved, may be larger than clinically acceptable in some cases. The results of this study show that the typically ignored expiratory data may be able to provide insight into patient condition when conventional methods fail. Clinically, these methods could have an impact in guiding MV therapy by providing clinicians with information about lung mechanics under the effect of patient SB effort.

Keywords: Respiration and ventilation, Biological and medical system modelling, Decision support systems
Abstract: Mechanical ventilation (MV) is the primary support for respiratory failure patients. MV treatment is difficult to manage due to the variable patient-specific disease state and response to MV treatment. Model-based methods have shown potential in providing unique, personalised information on patient condition for clinicians to guide MV treatment. This study presents a clinical observational trial to investigate the respiratory mechanics and quality of patient ventilator asynchrony conducted in a South-East Asia hospital intensive care unit (ICU). Pilot trial results included 7 study patients. Patient-specific respiratory mechanics Ers and Rrs were median 32.30 cmH2O/l [Interquartile range (IQR): 24.53-43.48] and 7.62 cmH2Os/l [IQR: 4.85-10.34]. The Normalised Area Under the Curve of Time-Varying Elastance (AUC-Edrs) across patients were 27.50 cmH2O/l [IQR: 24.97-28.34]. Patient ventilator interaction is assessed using an asynchrony index (AI), where AI had a median value of 27.0% [(IQR): 21.0-52.7]. Patient-specific respiratory mechanics displayed intra- and inter- patient variability, suggesting patients are different and evolved over time, and thus their MV settings should be patient-specific and time-varying. Model-based methods thus offer unique insight into patient-specific respiratory mechanics and the resulting opportunity to personalize and optimise MV based on evolving patient-specific needs.

Keywords: Respiration and ventilation, Biological and medical system modelling, System identification and validation
Abstract: Mechanical ventilation is a primary therapy for patients with acute respiratory failure. However, poorly selected ventilator settings can cause lung damage due to the heterogeneity of healthy and damaged alveoli. A commonly used lung protective strategy is titrating PEEP (positive end expiratory pressure) to minimum elastance. However, in clinical practice this initial PEEP is rarely changed. As lung disease evolves over time, the optimal PEEP required to maintain recruitment also changes. A predictive elastance model could be used to assess whether or not the current PEEP level should be changed based on whether a nearby PEEP had lower elastance. In this study, a physiologically relevant basis function model is developed and tested across the entirety of 18 recruitment manoeuvres. Accurate prediction of pressure and PIP to 10% across each of these data sets validates the functionality of this model.

Keywords: Respiration and ventilation, Biomedical imaging systems and image processing, Chronic therapy
Abstract: Electrical Impedance Tomography (EIT) is a non-invasive, radiation-free imaging method which can be utilized to measure regional lung ventilation distribution in patients with cystic fibrosis (CF). Up to now, the progression of CF-related lung diseases as well as therapy effects are monitored by spirometry or imaging methods such as computed tomography (CT). Since spirometry does not deliver any regional information of the lung function and CT cannot be applied frequently due to radiation, EIT might be a promising method to trace disease progression and therapy effects in CF-related lung diseases. In this preliminary study, the effects of physiotherapeutic breathing therapy on ventilation distribution in CF-related lung disease were investigated. Regional ratios of relative impedance changes corresponding to the forced expiratory volume in one second and the forced vital capacity (ΔI_FEV1/ΔI_FVC) were determined in three CF patients before and after therapy. Ventilation homogeneity was evaluated by a slightly modified version of the global inhomogeneity index (GI_FEV1/FVC). In all three CF patients, variations in regional ratios of ΔI_FEV1 and ΔI_FVC could be observed, whereby peripheral lung regions seemed to be more affected by the therapy than other lung regions. However, all three CF patients showed a more homogenous ventilation distribution after the physiotherapeutic breathing therapy than before. These first results indicate the suitability of EIT to trace physiotherapeutic breathing therapy effects in CF-related lung disease, which might benefit a more patient-specific therapy.

Keywords: Respiration and ventilation, System identification and validation, Biological and medical system modelling
Abstract: Abstract: Mechanical ventilation (MV) is widely used in Neonatal Intensive Care Unit (NICU) to support or fully control the breathing of patients of patients with respiratory distress syndrome (RDS). Current clinical practice involves clinician intuition or a generalized, “one size fits all” approach as inter-patient variability and lung heterogeneity can make selection of optimum ventilator settings difficult. Model-based methods can be used to identify patient-specific lung mechanics. This study aims to apply the single compartment lung model with an added endotracheal tube compensation term (ΔPETT) to fit neonatal clinical data to describe neonatal lung mechanics. Airway flow and pressure data was collected from 10 mechanically ventilated infants in the Christchurch women’s hospital NICU as part of an observational trial under standard care conditions. 205.9 hours of conventional ventilation (across 9 patients), and 53 hours of HFOV (across 3 patients) were recorded. The model fit was very good with median [interquartile range] of percentage MARD of 5.7 [5.2-6.3]% across all conventionally ventilated patients. Subgroup analyses was also performed based on weight, RDS and surgical cohort, and comparison between patients who were treated and not treated with surfactant therapy. Overall, the single compartment model fit was able to capture lung mechanics in premature infants, but showed large variability in mechanics across and within patients. Future work will use these results and models to better understand and monitor changes in patient condition for the improvement of delivered MV.

Keywords: Respiration and ventilation, System identification and validation, Biological and medical system modelling
Abstract: Mechanical ventilation is a core therapy for patients suffering from respiratory failure in an intensive care unit. However, incorrect ventilator settings can cause further lung damage due to the heterogeneity of lung diseases. A common lung protective strategy is titrating Positive end-expiratory pressure (PEEP) to minimum elastance during a staircase recruitment manoeuvre. Increases in applied PEEP can result in additional lung volume due to alveolar recruitment. However, excessive increases in pressure risks ventilator induced lung injury (VILI). Thus, a trade-off must be made between maximising recruitment and minimising risk. In this study, a non-invasive and clinically relevant model-based method to predict the quantity of this additional volume is developed and the accuracy assessed against measured volume changes. Initial results show reasonable accuracy in estimating gained volume with median [IQR] error of 40 [20 – 60] mL in PEEP changes of up to 9 cmH2O. Use of this parameter to predict lung mechanics at a higher PEEP level was also assessed and had an absolute median [IQR] error of 1.0 [0.5 – 1.8] cmH2O in predicting peak inspiratory pressure. These results offer an opportunity for clinicians to better optimise PEEP selection as patient condition evolves.

Keywords: Biological and medical system modelling, Cardiovascular system
Abstract: Curved blood vessels are the high incidence of cardiovascular and cerebrovascular diseases. The asymmetric nitric oxide (NO) distribution in blood vessels may maintain or regulate the curvature of blood vessels. To understand the distribution feature of nitric oxide in curved blood vessels, 3-D models were established to predict the distribution of NO in curved vessels. Animal experiments were also executed in vitro to verify the NO distribution. Simulation results show that the distribution of nitric oxide in the curved blood vessel is not uniform. The distribution of NO is different between the inner and outer sides of curved blood vessels. Immunohistochemical results showed that eNOS was mainly expressed in the vascular endothelium, and eNOS at the outside of the bend was significantly more than that at the inner side. This study is expected to open up new translational research to address vascular disease risk assessment and treatment.

Keywords: Biomedical imaging systems and image processing
Abstract: The dynamic SPECT reconstruction algorithm developed in our prior work reconstructs the parameters of the time activity curve for each image voxel directly from the projection images. In each iterations of the SPECT reconstruction beyond the static 3D MLEM (Maximum Likelihood Expectation Maximization) step, the algorithm performs a fitting process for each voxel in order to estimate the parameters describing the function of the examined organ considering that the time frames are not independent from each other. In real cases the fitted curve is nonlinear function of these parameters, it is usually described as the sum of exponential functions. In order to estimate the parameters properly, an iterative root-finding method is applied. In the current study the Newton-Raphson method is used. The selection of a proper initial value for the root-finding method is critical in order to achieve convergence of the fitting process. If the initial guess is not appropriate, the rootfinding algorithm can diverge or converge to an inappropriate parameter set that can result in unacceptable reconstructed parameters. This affects then the subsequent MLEM iterations, also neighboring voxels and breaks the reconstruction. In this work we investigated different methods to calculate the initial values of the fitting process and evaluated the reconstructed parameter set of the dynamic SPECT reconstruction algorithm. Three different methods are investigated, one that uses the fitted parameters of the previous MLEM iteration, one that is based on the sum of the geometrical series of the exponentials and one that calculates the best guess using both methods. The three methods were compared by benchmark reconstruction cases using a mathematical phantom. In each reconstruction different initial value selection method was applied then the time activity curves of the voxels belonging to the same tissue were statistically evaluated using the reconstructed parameters. In the study no significant differences were found in the mean value of the reconstructed parameters. The standard deviation of the parameters was similar between the two simple approaches, however, the combination of the methods resulted in better statistical performance.

Keywords: Biomedical imaging systems and image processing
Abstract: In the past, since the reperfusion part is less than 10% of the total vascular system composed of multiple inputs, such as lungs, the gamma-variate function was used as a method of eliminating the reperfusion component. In this research, an analytical method that separates and evaluates all blood flow systems by modeling the impulse response of the blood flow system with multiple inputs was proposed. In addition, a calibration method to adjust the input/output waveform areas was also proposed. The proposed method was verified in a patient with lung cancer. The lung total blood volume using the contrast MR image and the global lung blood volume estimated from the pulmonary artery and the signal strength of the left atrium were also compared. Moreover, the lung field reperfusion percentage was compared with the cardiac output detected by Fast GE Cine.

Keywords: Biomedical imaging systems and image processing
Abstract: Ultrasound Tomography (UST) and Electrical Impedance Tomography (EIT) images can be computed using Baye's inference. Prior information is required for the solution of these two inverse problems. In order to build prior information of the human chest, the use of already taken CT-Scans is recommended because radiation exposure will be avoided. CT-Scans are taken with the arms above the head. UST and EIT images are taken with the arms below the head. The present work shows that the movement of the scapula intersects the plane of transducers of the UST or the plane of electrodes in EIT. It is necessary to correct the position of the scapula before the use of a CT-Scan to build an anatomical atlas for lung monitoring.

Keywords: Simulation and visualization, Cardiovascular system, Biomedical imaging systems and image processing
Abstract: The valve-sparing aortic root surgery is a common treatment of several aortic diseases. Although, the aortic valves are generally fixed in an even distribution during these operations, recently developed special tools enable the back-sewing of the valves in the original, patient specific distribution that likely provides better hemodynamic functions after the operation. The aim of the presented work is to quantitatively analyze the hemodynamic outcomes of the alternative valve-sparing root surgery operations by using computer simulation. In the paper the developed modeling, elastic, and fluid simulation framework of the aortic root is introduced. The workflow implemented by the framework covers the patient specific aortic root geometry modeling using CT or MR data, finite element simulation based virtual aortic valve sewing, and blood flow parameter calculation in the aortic root. Our custom-developed simulation techniques use simplifications applicable directly for aortic valve simulation resulting in reasonable computational complexity. The preliminary results of aortic valve modeling of 10 patients are presented at the end of the paper.

Keywords: Biomedical imaging systems and image processing
Abstract: This study develops a method to identify equivalent viscous damping for breast cancer diagnosis in a Digital Image Elasto Tomography (DIET) system. Displacement data of over 14,000 reference points on the breast surface of 7 breasts (4 cancerous and 3 healthy) from 4 patients were captured using the DIET system. An ellipse fit was utilized to calculate the work done and consequent viscous damping constant for each reference point. The gradient of the empirical cumulative distribution (CDF) of the viscous damping for each breast was calculated to detect and localize tumor segments by CDF variability. The gradient of the empirical cumulative distribution exhibited a flatter section early on in the tumor segments followed by a rise in gradient when compared with healthy tissue. A gradient ratio was defined to quantify this hypothesis, with good proof of concept accuracy for diagnosis and location of the tumor matching original mammography report. The overall results show the potential of identifying viscous damping for breast cancer diagnostics in a DIET system.

Keywords: Biological and medical system modelling, Diabetes management, Biosignal analysis, processing and interpretation
Abstract: This paper addresses the problem of mathematical deconvolution for the estimation of unknown inputs in linear discrete-time state-space models. We apply our deconvolution algorithm to the modeling of blood glucose (BG) concentration for people with type 1 diabetes (T1D). We present a method using an activity tracking watch, a continuous glucose monitor and an insulin pump to study the effect of physical activity on BG concentration for people with T1D. The physical activity signatures are represented by an unknown input, also referred to as a "net effect". In addition, the net effect captures the unmodelled BG variations, eg. mismatches in meal estimation, and circadian metabolic variations. We test our method using data from a clinical study. We show the glucose net effect traces associated to physical activity for a specific patient during 20 consecutive days, and the glucose net effect traces associated to physical activity for eight subjects under identical conditions. The net effect signatures can be used to (i) reproduce experiments with different insulin administration strategies, (ii) build a physiological model of glucose-insulin dynamics able to simulate a physical activity in people with T1D, and (iii) design a model-based control algorithm able to predict the effect of physical activity on the blood glucose concentration.

Keywords: Biosignal analysis, processing and interpretation, Medical technology, Metabolic system
Abstract: Neonatal hypoglycaemia is common in at-risk infants and can cause adverse neurologic outcomes in later life. Continuous glucose monitoring (CGM) technology offers a way to continuously monitor patient condition, helping to detect hypoglycaemia as well as provide insight into the general glycaemic state of the patient. Characterising Glycaemic States can be easily done by eye, but no simple, clinically relevant algorithm exists to do this characterisation analytically or computationally. This paper presents such an algorithm to characterise Glycaemic States and detect State Changes. This algorithm was developed on a cohort of 366 infants, using a total of 12356 hours of CGM sensor data. State Changes were defined as an intersection between a 6-hour rolling average of the CGM trace and the average of the whole interstitial glucose CGM trace, with a 5 hour minimum crossover threshold defining a single State. The majority of infants were found to have experienced less than 2 State Changes in the first 48 hours of birth (279 of 366 patients, 76%). The median number of State Changes per day was 0.68 [IQR: 0.60, 1.14], while the median absolute change in IG over a State Change was 0.6 mmol/L [IQR: 0.4, 0.9 mmol/L]. Visually, the majority of algorithmically characterised State Changes matched CGM traces characterised by eye. Future use of the algorithm could associate the State Changes with clinical outcomes.

Keywords: Neuro-prosthetics, Biosignal analysis, processing and interpretation, Control systems
Abstract: Control systems for human physiotherapy exercises based on functional electrical stimulation (FES) have provided excellent performance in several setups. Myo-controlled neuroprostheses use electromyography (EMG) for timing and intensity control of stimulation applied to these exercises, estimating not only the volitional activity (from the patient) but also the evoked activity (from FES). A typical EMG response to FES starts with the stimulation artifact, followed by an excitation curve called M-wave. To extract volitional and evoked components, we first need to find the inter-pulse-intervals (IPIs), i.e., the EMG signal between stimulation artifacts. For that, we have developed a method for two-channel stimulation artifact detection for EMG signals which are not hardware-synchronized to a FES stimulator. First, the artifact detection approach marks all potential artifacts based on one of three adaptive threshold-based detection methods (mean/standard deviation, median/MAD and quantiles). Subsequently, for IPI extraction we cluster the potential stimulation artifacts to cross-correlate the resulting potential stimulation artifact vector with a vector of expected artifacts based on the stimulation and EMG frequencies. For evaluation, we performed tests on two benchmark datasets obtained from FES-assisted walking with two hardware setups. We found more than 95% success rate for both hardware setups using the adaptive threshold method independently on the selected method for choosing the threshold. Because of its low computational demands, we recommend the mean/standard deviation approach.

Keywords: Biological and medical system modelling, Diabetes management, System identification and validation
Abstract: We present a novel approach to the selection of dynamic system models with regard to stabilisability in terms of their observer error dynamics. For that, we introduce the gap-metric for the observer error dynamics as a distance measure. Models are selected for different subsets in such way, that the gap-metric is minimised. The procedure is subsequently applied to classify parametrised G"ottingen Minipig models that were obtained from animal experimental data, into different subsets. These obtained classes are then set up as the basis for an optimal experimental design procedure and result in improved convergence properties, as well as, in reduced implementation effort.

Keywords: Biological and medical system modelling, Neurosystems, Simulation and visualization
Abstract: A three-dimensional model of the aqueduct of Sylvius (AS) is used to simulate the behavior of pressure and velocity of cerebrospinal fluid (CSF). Magnetic resonance imaging (MRI) of the brain, obtained from databases were used for the reconstruction of the model. Boundary conditions from different authors were use in order to prove the proposed model.

Computational Fluid Dynamics (CFD) tools predicted that the maximal velocity occurs at the narrowest point of the AS. On the other hand, the inclusion of the third and fourth ventricles in the models is not significant on both pressure drop and velocity profiles. This work allows to analize variations in pressure and velocity in the AS with different boundary conditions.

Keywords: Biological and medical system modelling, Respiration and ventilation, Quantification of physiological parameters for diagnosis assessment
Abstract: Specific pulmonary compliance is a widely used metric in neonatal mechanical ventilation, capturing the intrinsic ability of the lung to expand with pressure normalized by the functional residual capacity of the lung. Previous studies have found specific compliance to be similar for both neonates and adults. However, it is unclear what this implies about the structure and growth of the lung. To model the factors affecting specific compliance, a simple model is developed separating respiratory compliance into material-dependent and volume-dependent components. It provides reasonable estimates of compliance for experimental clinical data. The model demonstrates specific compliance is dependent on the alveolar tissue elastic modulus, Poisson’s ratio, septal thickness to alveolar radius ratio, and surface tension effects. The model also indicates specific compliance may not be entirely independent to lung volume as the lung grows, due to surface tension effects varying with alveolar diameter, and the possibility of a varying alveolar wall thickness to diameter ratio. To experimentally evaluate trends in specific compliance, dta from 9 infants and 8 adult mechanical ventilation patients was recorded. Median compliance values were 0.51 mL/cmH2O/kg for adult patients and 0.78 mL/cmH2O/kg for neonatal patients (p = 0.098). This result demonstrates mechanically ventilated neonates and adults have statistically similar values of specific compliance, in agreement with prior studies. Treating specific compliance as a constant and using a least-squares fit, the trend of neonatal compliance increasing with body weight can be used to predict typical ranges of adult compliance values, indicating the non-linearity of compliance varying with volume is likely not significant in reality, supporting the use of specific compliance as a measure of the underlying tissue distensibility. However, the variation in values for each cohort suggest that specific compliance, like lung elastance, can vary significantly, and while overall trends hold, central values are not necessaril indicative. Overall, the model provides a unique analysis and insight using a simplified, mechanics relevant model, which is could be generalized to other studies