Abstracts Track 2026


Nr: 30
Title:

Tailoring Neuromuscular Dynamics: A Modeling Framework for Realistic sEMG Simulation

Authors:

Alvaro Costa Garcia

Abstract: Surface electromyography (sEMG) is a fundamental tool for analyzing muscle activity, yet its signal interpretation can be affected by multiple biomechanical factors. This study presents an advanced computational model to simulate sEMG signals during muscle contractions, integrating five key components that replicate the chain of processes from motor intent to voltage variations on the skin. These components include the motor control system, motor neurons, muscle fibers, biological tissues, and electrodes. sEMG signals were simulated for isotonic and isometric contractions under two force conditions and compared with experimental data obtained from elbow flexion tasks. Results showed a high level of similarity between simulated and real signals, capturing both temporal and spectral characteristics. Moreover, the study revealed a correlation between the distribution of muscle fiber types and changes in the spectral distribution of the simulated signals.Potential applications of this framework include improved interpretation of sEMG in clinical and sports settings, non-invasive estimation of muscle fiber composition, simulation of coordinated multi-muscle activations and kinematics, and supporting early detection and monitoring of muscular disorders, as well as personalized training strategies.

Nr: 74
Title:

Developing a Robust, Programmable Stenotic Flow Phantom for High Frame Rate Ultrasound Imaging

Authors:

Yi Zheng, Tomas Jansson, Adrian Thomson, Tom MacGillivray, Marcus Ingram, Jan D'Hooge and Carmel Moran

Abstract: Background, Motivation and Objective Atherosclerotic diseases are one of the leading causes of death worldwide. As disease progresses, the presence of fatty plaques affects regional hemodynamics behaviour. Wall shear stress (WSS), a key parameter for assessing both the development and prognosis of disease, can be captured by high-frame-rate (HFR) ultrasound, for example through unfocused beam transmissions coupled with coherent compounding to compensate for degraded spatial resolution. In-vitro flow phantoms, made from tissue-mimicking materials to recreate clinically-relevant blood flow, is an effective experimental tool for developing and testing new ultrasound technologies and methodologies. The aim of this study was to derive trajectories of near-wall microbubbles (MBs) using a programmable, stenotic, flow phantom, and analyze the HFR phantom images acquired using an ultrasound advanced open platform (ULA-OP) in order to derive near-wall flow velocities and hence wall shear stress. Methods PVA cryogels, prepared with different freezing-thawing cycles, were acoustically and mechanically characterized to select the most suitable vessel-mimicking material for this study. The phantom was constructed by casting PVA cryogel in 3D printed rods (diameter: 3 mm) of 0% to 80% stenoses, validated using a preclinical ultrasound scanner (Vevo3100®) at 40 MHz. These phantoms of different levels of stenosis represent healthy (0%), mild (< 50 %), moderate (50% -70%), and severe (>70%) clinical scenarios. Programmable flows were generated by a gear pump, via an external closed-loop micro-controller. The system step response was derived from the voltage data tracked by a digital oscilloscope. Both steady state and physiological flows were designed for system calibration and clinical relevance. To validate this phantom, homogenized sparse MBs (SonoVue®) were introduced to the stenotic phantom at pre-calibrated steady and physiological flow rates. Three plane waves (-60, 00, 6 0) were transmitted using a linear 7.5 MHz transducer on an ULA-OP (256 channels) to acquire high frame rate B-mode scans at 500 fps after compounding. Lumen boundaries were dynamically segmented to track the wall displacements in each mimicked cardiac cycle. Ultrasound localization microscopy (ULM) was used to derive near-wall flow trajectories, and a Savitsky-Golay filter was used to interpolate a smooth velocity map while keeping the flow features. Finally, WSS was derived from the velocity gradient. Results/Discussion PVA cryogel with 2-3 freezing-thawing cycles demonstrated balanced acoustic and mechanical potentials as a vessel-mimicking material. The manufacturing pipeline in this study is effective for manufacture of a robust, programmable flow phantom. Using a high-resolution ultrasound scanner, flow jets and regional recirculation were clearly observed at throat and post-stenosis areas, respectively. Using our algorithm, in phantoms mimicking a healthy vessel and moderate stenosis, the differences between simulated and measured velocities and WSS were 1.7% at centerline and 6% near wall, respectively. In future studies, for more robust WSS measurement, turbulent flow components will be implemented to compensate for disturbed flow trajectories.

Nr: 339
Title:

Nanomaterials for Diagnostics, Therapeutics and Preventing Spread of Infectious Diseases

Authors:

Tohid Didar

Abstract: The biological/non-biological interface system is an important cornerstone for the fabrication of a wide range of biomedical devices. Platforms as diverse as lab-on-chip and point-of-care diagnostics, 3D tissue culture scaffolds, organs-on-chips, implants and antimicrobial surfaces all rely on the effective interaction of cells and/or bio-recognition elements (proteins/peptides, enzymes, oligonucleotides, etc.) with non-biological surfaces. Design and engineering of micro/nano patterned interfaces provide powerful tools to study biological phenomena at micro and nano scale and to develop novel technologies for diagnostics, therapeutics and preventing the spread of infectious diseases. I will present an overview of our research on design and development of transformative nanomaterials and their integration into in vitro systems such as lab on chip devices, flexible sensing interfaces, smart food packaging and antimicrobial and pathogen repellant coatings as well as in vivo applications to develop efficient medical devices such as injectable hydrogels, anti-thrombogenic and antimicrobial catheters, vascular grafts and implants.

Nr: 341
Title:

An Enhanced CGR Based Representation of DNA Sequences and Application of Digital Signal Transform Techniques in Sequence Analysis

Authors:

Sanatan Das

Abstract: Determining biological sequence similarity is pivotal for understanding evolutionary relationships among species, gene function analysis, predicting gene and protein functions, and finding relevant sequences in databases. In the last few decades, the researchers have developed different methods of the digital encoding of the DNA sequences and a number of alignment-free sequence analysis techniques for extracting the inner patterns from DNA sequences. It is already proved in many studies that, the alignment-based sequence similarity analysis consumes high memory space and runtime and cannot be computed using normal computers for the modern species sequences. In this research, we have proposed the use of Digital Signal Transform techniques for DNA sequence analysis. First, we propose an information theory-based method for determining k−mer value using entropy maximization for a collection of sequences of varying lengths. Then, we propose an enhanced count based CGR algorithm for converting the k−mer strings to numerical data. Then, we apply multiple Digital Signal Transform methods for decomposing the signals and extracting the inner patterns of sequences. In this research, we used Fast Fourier Transform (real and imaginary coefficients), Haar and Daubechies 4 (db4) Wavelet Transform for feature extraction. In case of Haar and db4 Wavelet Transform, we used the approximation coefficients to create 2D feature vectors. Then, we used the ’cosine’ distance for calculating distance matrix and Neighborhood-Join clustering for phylogenetic tree construction. We applied these methods on two benchmark datasets (25 Cichild Fish and 8 Yersinia Strains) and received the highest scores. We have performed a comparative time and memory (feature vector size) analysis of these Digital Signal Transform techniques. The comparative results show that all three methods are effective in extracting the inner patterns of the DNA sequences, however, Fast Fourier Transform shows best results and achieves the highest score for multiple k−mer values on these datasets. Haar Wavelet Transform reduces the dimension of the feature vectors and runs faster than other methods. Our experiment and findings demonstrate the potential applicability of Digital Signal Processing techniques for sequence similarity analysis.

Nr: 343
Title:

Variable-Based Calibration Evaluation Methods for Binary Outcome

Authors:

Hiroe Seto, Shuji Kitora, Asuka Oyama, Hiroshi Toki, Ryohei Yamamoto and Michio Yamamoto

Abstract: When developing a probabilistic prediction model to predict future disease risk, good calibration is required for subgroups with high demand for the model, such as obesity and the elderly. However, existing calibration evaluation methods cannot evaluate whether the developed prediction model returns good prediction values for each subgroup. This is because existing evaluation methods evaluate calibration based on predicted probability, and cannot evaluate calibration for variables of interest, such as weight and age. Furthermore, existing calibration evaluation methods often fail to detect model bias. Therefore, in this study, we developed two methods for evaluating calibration for variables of interest. One is the variable-based probabilistic calibration error (VPCE), which allows for quantitative evaluation, and the other is the variable-based probabilistic calibration plot (VPC-Plot), which allows for visual evaluation. We then investigated the characteristics of the proposed method and its validity as an evaluation method through mathematical proof and numerical experiments. Furthermore, to demonstrate its usefulness in real-world data analysis, we evaluate the calibration of the developed diabetes prediction model to variables using the National Health Insurance Database (Kokuho Database; KDB) in which approximately 2 million insured individuals are included every year.

Nr: 358
Title:

A Low-Cost Non-Invasive Flexible Sensor Matrix Patch for Neonatal Respiratory Monitoring

Authors:

Zhanglei Jin, Eugen Koch, Haochen Xu, Johannes Busse and Andreas Dietzel

Abstract: Respiratory complications remain a leading cause of mortality in neonates and preterm infants, particularly in resource-limited regions. For infants with underdeveloped lungs, real-time monitoring is critical for detecting respiratory distress and coordinating timely ventilatory support. This study presents a low-cost, non-invasive flexible sensor matrix patch for neonatal respiratory monitoring. The system employs a hybrid fabrication process that combines custom double-sided printed carbon resistors with commercial flexible printed circuit board (FPCB) technology with 18 µm copper traces deposited built on a 25 µm polyimide (PI) substrates. 15 µm thick adhesive PI layers are laminated as cover layers on both sides, protecting the circuit while leaving open windows at sensing sites for subsequent carbon printing. A two-layer technology with copper interconnects provides high routing resolution together with mechanical robustness, enabling high-density and scalable sensor manufacturing. The sensor array comprises symmetrical carbon resistors printed on both sides at corresponding positions interconnected through copper traces and vias in low-strain regions ensuring mechanical stability. 14 measurement units base on voltage divider are compactly distributed over a length of approximately 60 mm, sufficient to capture abdominal deformation of neonates and thus to detect respiratory motion. The symmetric architecture not only enhances compactness but provides also passive thermal compensation. The sensor matrix geometry is designed for stretchability and compliance, allowing close conformity to abdominal motion. This reduces skin–sensor interface stress from breathing or movement, and improves comfort and signal stability. Experimental results under static bending with 20, 40, 60, 80 mm radius yielded the gauge factor of 7.4, indicating high strain sensitivity. Hysteresis was below 10% of full-scale output, demonstrating acceptable linearity. To assess long-term stability, bending tests up to 10,000 cycles were performed, during which no channel failure occurred and the sensor array maintained functional signal output throughout the test duration. Sensor signals were acquired at approximately 57 Hz using a custom PCB based on the PSoC platform, with data transferred via USB for real-time visualization and processing. To validate the sensor arrays before clinical use, tests on volunteers monitoring abdominal breathing have been carried out. In summary, this stretchable sensor matrix, fabricated using a hybrid double-sided process, demonstrates promising performance. It offers a low-cost, comfortable, and non-invasive solution for real-time neonatal respiratory monitoring and has strong potential for integration into neonatal care platforms and future intelligent monitoring systems.

Nr: 416
Title:

Fusion of Light-Sheet Autofluorescence Microscopy and Histology in Tissue Punches Using LitSHi

Authors:

Marcel Brettmacher, Philipp Nolte, Diana Pinkert-Leetsch, Felix Bremmer, Jeannine Missbach-Guentner and Christoph Rußmann

Abstract: Histological assessment is a cornerstone of tissue diagnostics but is inherently limited to two-dimensional (2D) sections, providing only partial insight into the three-dimensional (3D) organization of tissue. Light Sheet Fluorescence Microscopy (LSFM) enables volumetric imaging of intact samples and offers valuable spatial context, yet its integration with histology remains challenging due to differences in resolution, orientation, and sectioning-related deformations. In particular, aligning 2D histological sections within 3D LSFM volumes is computationally demanding and often requires manual intervention or expert knowledge. We present LitSHi (Light Sheet meets Histology), an automated registration tool for fusing LSFM data with histological sections of tissue punches. LitSHi performs 2D-to-3D alignment fully automatically, eliminating the need for fiducial markers or manual pre-alignment. The tool enables efficient and reproducible multimodal image fusion and is readily applicable in routine workflows. LitSHi was evaluated on tissue punch samples from testicular tumors and pancreatic tissue, where it achieved accurate structural correspondence between LSFM and histology. Compared to manual and semi-automated approaches, LitSHi significantly improved both alignment accuracy and processing efficiency. By simplifying multimodal tissue integration, LitSHi supports practical applications in digital pathology and provides a foundation for advanced computational and AI-based tissue analysis.

Nr: 417
Title:

Estimation of Rheological Properties of Red Blood Cells in Patients with Chronic Lymphocytic Leukemia Using a Microfluidic System and Developed Software Image Flow Analysis: The Impact of Targeted Therapies

Authors:

Anika Alexandrova-Watanabe, Emilia Abadjieva, Tihomir Tiankov, Aleksandar Iliev, Ariana Langari, Miroslava Ivanova, Lidia Gartcheva, Sashka Krumova, Dimitar Trifonov and Svetla Todinova

Abstract: Chronic lymphocytic leukaemia (CLL) is characterised by the progressive accumulation of abnormal B lymphocytes and remains incurable using standard therapies. Although targeted treatments have been developed, little is known about their impact on red blood cell (RBC) rheology. The study aims to evaluate the aggregation and deformability of RBCs in CLL patients compared to healthy individuals and to explore the hemorheological effects of the commonly applied therapeutics Obinutuzumab/Venetoclax and Ibrutinib. using a microfluidic system – BioFlux, along with newly developed Software Image Flow Analysis. Using the BioFlux microfluidic system, the RBC aggregation is studied over a range of shear rates from 0 to 445 s⁻¹. High-resolution images are obtained during the flow of RBC suspensions through microfluidic channels (75x350 μm) of a 24-well 0-20 dyn/cm2 plate. These images are analyzed by a novel algorithm developed to evaluate the RBC aggregates, and a software program created based on this algorithm (Software Image Flow Analysis). The program, written in the ImageJ Macro Language, follows a structured workflow to analyze RBC aggregates data and provide software visualizations based on the area of the aggregates. For complex analysis, parallel with the evaluation of RBC aggregation is quantitatively assessed the deformability of RBCs under dynamic flow by the BioFlux microfluidic system. For this reason, another new software application has been developed to study the deformability of RBCs in flow using LabVIEW and MATLAB platforms. With the help of the elaborated microfluidic software image analysis is determined the elongation index (EI) of RBCs in the range of shear rates (89–625 s⁻¹). The results show that the treatment with a combination of Obinutuzumab/Venetoclax does not affect RBC aggregation. In contrast, targeted therapy with Ibrutinib preserves RBC aggregation in CLL to levels observed in healthy controls, demonstrating that Ibrutinib attenuates the CLL-association alterations in RBC rheological properties. RBCs from untreated CLL patients exhibit significantly lower EI values at all shear rates, indicating poor deformability. RBCs from patients treated with Ibrutinib - show substantial recovery of deformability, with EI values approaching those of healthy controls at lower and intermediate shear rates. However, at high shear rates (535–625 s⁻¹), EI remains significantly reduced, suggesting incomplete recovery of membrane flexibility. Our findings demonstrate that CLL is associated with altered RBC aggregation and deformability, and these rheological properties are differentially affected by targeted therapies. These are critical for predicting potential treatment-related complications and side effects, especially with respect to blood flow dynamics, which could be assessed in vitro using a microfluidic device. Acknowledgements: This work is realised in the Center of Competence for Mechatronics and Clean Technologies “Mechatronics, Innovation, Robotics, Automation and Clean Technologies” – MIRACle, with the financial support of contract No. BG16RFPR002-1.014-0019-C01, funded by the European Regional Development Fund (ERDF) through the Programme “Research, Innovation and Digitalisation for Smart Transformation” (PRIDST) 2021–2027. This work is also supported by Grant KP-06-H73/3, competition for financial support for basic research projects—2023, Bulgarian National Science Fund.

Nr: 423
Title:

Beyond Static Models: What Is the Optimal Framework for Evaluating Dynamic Functional Connectivity?

Authors:

Stefania Coelli, Martina Corda and Anna Maria M. Bianchi

Abstract: Brain functions depend on the synchronized activity across space and time, which facilitates information transfer between functionally linked cortical regions (i.e., regions with statistically interdependent activity). We can measure the strength of these functional links using various techniques that assess these relationships across time, frequency, or information domains. However, these interactions are not static; they change dynamically over temporal scales from slow hemodynamic responses up to fast electrophysiological dynamics (like time-locked evoked potential and oscillations). Thanks to the continuous improvements in neuroimaging, computational power, and analytical methods, a major focus in recent research has been the analysis of dynamic functional connectivity (dFC), thereby moving beyond static models and restoring the crucial time dimension to our understanding of brain organization [1]. In particular, instantaneous (sample by sample) dFC methods are required to track rapid changes in brain connectivity, as fast neural interactions, evolving on sub-second timescales, cannot be captured by other approaches, such as those relying on sliding windows. Two categories of instantaneous dFC techniques exist: instantaneous phase-based methods and time-varying multivariate autoregressive (tvMVAR) models. In this study, a methodical and numerical performance analysis of tvMVAR properties and an evaluation of how these models respond to connectivity changes is provided to improve the understanding of correct methods application and results interpretation. The development of a comprehensive evaluation framework was essential for the assessment and comparison of the performance under both controlled and real conditions. TvMVAR models, adapted using the classical Kalman filter (CKL) and the generalised linear Kalman filter (GLKF), were evaluated using simulated datasets with known ground-truth. The most effective method, as determined by quantitative analysis, was then applied to evoked potentials from a benchmark dataset, which is known for its well-established connectivity patterns. The developed assessment framework offered practical guidelines for working with real data, emphasising the importance of selecting the correct method and optimizing the needed parameters based on the dataset under analysis. To conclude, an application of the derived practical indications was performed to analyse electroencephalography (EEG) data acquired on 13 healthy subjects during a motor task, to explore the dFC patterns underlying motor control. Results demonstrated fast task-related modulations of the analysed network, specifically highlighting the dominant role of the contralateral motor cortex as a source of information flow [2]. We conclude that dFC represents a powerful tool for exploring neural signal interactions and network reconfigurations over time, enabling the identification of specific synchronization patterns among activity sources and enhancing the understanding of complex time-varying neural mechanisms. References: [1] Coelli, S. et al.,"The time-varying brain: a comprehensive review of dynamic functional connectivity analysis in EEG and MEG." JOURNAL OF NEURAL ENGINEERING 22.5 (2025) [2] Corda, M. et al. "Tracking Dynamic Functional Connectivity Using Time-Varying Multivariate Autoregressive Models: an EEG Study of Sensorimotor Processes." IEEE Engineering in Medicine and Biology Society. 2025.

Nr: 425
Title:

High Users Management with AI: An Integrated Smart Solution to Improve Care and Optimize Resources for Frequent Emergency Users

Authors:

Teresa Bastos Lopes, Mariana Dias, Marta Lopes, Luís Silva, Ana Januário, Andreia Mesquita, Carla Menino, Carla Vidinha, Fausto Honoré Silva, Mafalda Gonçalves, Maria José Guimarães, Pedro Casimiro, Rui Malha, Ana Londral and Federico Guede-Fernandez

Abstract: Overcrowding in emergency departments (ED) is a global public health problem with a negative impact on the quality of care provided, service efficiency, and resource utilisation. A small subgroup of patients, referred to as high users (HU), comprise approximately 3.5% to 29% of ED users yet account for more than 60% of all ED visits, and are characterised by multimorbidity, complex care needs, and a high psychosocial burden. At Garcia da Orta Hospital, in Portugal, this reality led to the implementation of the Grupo de Resolução dos High Users (GRHU), a multidisciplinary team dedicated to the case management of HU of the ED. This programme integrates individualised multidisciplinary consultations (MC), with the aim of reducing ED utilisation and improving health outcomes. However,GRHU faces challenges in selecting patients for intervention, particularly in identifying those for whom the intervention is likely to have the greatest impact. This study aims to identify patient profiles based on their response to the GRHU intervention and to develop predictive classification models to support patient selection. Clinical and sociodemographic data from the hospital database over three years (2022–2025) were used, covering 8250 patients, of whom 256 underwent the GRHU intervention. Patient profiles capturing the level of impact of the GRHU intervention were identified using unsupervised analysis methods, including K-means, Ward’s hierarchical clustering, and Gaussian mixture models: K-means for identifying compact utilisation patterns, Ward’s method for variance-based and hierarchically interpretable groupings, and Gaussian mixture models for modelling heterogeneous and overlapping patient response profiles. The analysis used clinically defined impact metrics, including ED visits, specialist consultations, hospitalisations, and MCs, measured over periods of 3, 6, and 12 months around the intervention. The clustering approach that best identified distinct response profiles was Ward’s hierarchical method with three clusters. The model used a 6-month window and the following variables: number of ED visits after the first MC, reduction in ED visits following the first MC, time between the first and last MC, and number of MC. This analysis was applied to the 145 patients with complete data within the 6-month window. This solution was selected based on internal validation metrics (Silhouette=0.55; Davies–Bouldin=0.94; Calinski–Harabasz=93.25) complemented by external criteria of clinical interpretability, reflecting clear and consistent patterns in changes in ED utilisation. Three distinct profiles were identified: a profile characterised by a marked reduction in ED visits shortly after the first MC, potentially indicative of a strong response to the intervention (n=106); a second profile associated with a longer intervention duration, followed by a reduction in ED visits (n=24); and a third profile displaying more heterogeneous trajectories, with limited or no reduction and in some cases an increase in ED visits(n=15). As future work, based on these profiles, supervised classification models will be developed to predict each patient’s response profile based exclusively on pre-intervention variables. This way, the team can predict the level of impact that the intervention will have on the patient, supporting clinical decision-making and contributing to a more efficient allocation of resources and optimisation of the GRHU intervention.

Nr: 429
Title:

Thermal Guardian: An Assistive Wearable System for Home-Based Monitoring and Prevention of Bathing Related Health Risks in Older Adults

Authors:

Joseph Arthur Koo, Ken Argani Toendan, Ethan Taku Shimada, A. K. M. Akaiduzzaman and Zilu Liang

Abstract: As populations age globally, enabling older adults to live safely and independently at home has become a central challenge for healthcare systems. Many serious health incidents among older adults occur not in hospitals, but during routine daily activities carried out in private residential settings, where continuous supervision is unavailable. Bathing-related accidents are a major public health concern. In Japan, approximately 19,000 bathing-related deaths reported annually. Bathing, especially during winter, exposes older adults to large temperature gradients that can trigger abrupt physiological responses. Experimental studies have shown that exposure to cold dressing rooms (10–15 °C) induces significantly greater systolic blood pressure fluctuations in elderly individuals compared to younger populations. Subsequent hot water immersion can further alter hemodynamic and autonomic regulation; reduced sympathetic tone following immersion may lead to hypotensive syncope minutes later, increasing the risk of silent drowning or collapse. Because these events often occur without immediate awareness or witnesses, delayed intervention frequently results in severe outcomes. In this work, we present an ongoing project, Thermal Guardian, an assistive wearable system designed to enhance bathing safety for older adults through continuous physiological and environmental monitoring. By integrating real-time sensing, adaptive safety thresholds, and immediate user feedback within a smart home context, the proposed system aims to support safer independent living and reduce the burden of preventable home-based health emergencies. Thermal Guardian is designed as a preventive assistive technology that continuously monitors both physiological status and environmental conditions throughout the bathing process. From a clinical perspective, the system targets early indicators of cardiovascular and autonomic instability, conditions that frequently precede syncope or collapse but are rarely detected in home settings. As shown in the figure, the system consists of two coordinated hardware components: (1) a wearable arm-mounted device and (2) a complementary ambient IoT sensor placed in the bathing environment. This dual-layer design enables contextual interpretation of physiological changes, distinguishing benign fluctuations from clinically relevant risk escalation driven by environmental stressors. The wearable device captures key physiological signals relevant to bathing-related risk, including heart rate, oxygen saturation, skin temperature, and movement patterns. These signals are complemented by environmental measurements of ambient temperature and humidity, which are known to modulate cardiovascular and autonomic responses during bathing. Collected data are transmitted to a companion Android application for real-time processing and visualization. When patterns indicative of elevated risk are detected, the system provides immediate auditory and haptic feedback to prompt user awareness and corrective action. Clinically, this tiered intervention strategy aligns with established principles of preventive care and patient safety, emphasizing early warning and timely intervention before critical events occur.

Nr: 430
Title:

Wearable ECG and Temple-PPG System with Stress-Task Dataset for Long-Term Cuffless Blood Pressure Monitoring

Authors:

Yiyao Feng, Chijing Wang, Yuki Ban, Toshihiro Takahashi, Takashi Inoue, Masanori Hiramitsu and Shin'ichi Warisawa

Abstract: Introduction: High blood pressure (BP) is a major cardiovascular risk factor, and understanding its daily-life variability is essential for detecting masked hypertension and dietary factors. This requires continuous BP monitoring under real-life conditions. However, ambulatory blood pressure monitoring (ABPM) relies on repeated cuff inflation and sparse measurements, causing discomfort and interference with daily activities, which limits its suitability for long-term monitoring. To address these limitations, previous studies have explored cuffless BP estimation using electrocardiography (ECG) and photoplethysmography (PPG) in laboratory settings. However, most systems employ finger- or ear-based PPG sensors that hinder routine activities or cause discomfort during prolonged use. Thus, the main contributions of this study are as follows. First, to enable comfortable long-term cuffless BP monitoring in daily life, we developed a wearable vest system for synchronized ECG and temple-PPG acquisition, enabling continuous BP trend analysis during multi-day monitoring. Second, to overcome the limited variability of resting-state BP datasets, we introduced stress tasks to induce BP fluctuations. The resulting data captures BP dynamics from baseline through stress and recovery, providing a broader dynamic range relevant to real-life conditions and a valuable basis for future cuffless BP modeling. System Design and Experiment: The wearable system consists of a vest-based platform for ECG acquisition and a temple-mounted PPG sensor, with sampling at 1000 Hz. A standard three-lead ECG was used, with electrodes placed on the bilateral clavicle regions and the iliac crest. A vest form factor was selected to stably fix ECG electrodes on the upper body while enabling prolonged wear during daily activities. To enhance wearability and reduce cable motion, the vest has a double-layer structure that secures and conceals sensor wiring. For PPG acquisition, the sensor was placed at the temple, which allows reliable detection of superficial temporal artery pulse signals without occupying the hands or introducing discomfort commonly associated with ear- or limb-based placements. To obtain physiologically diverse data for cuffless BP model development, we collected measurements under multiple physiological states rather than relying on resting conditions only. We conducted a 1.5-hour experiment with 28 participants involving rest, light walking, and structured cognitive stress tasks—including Stroop test, OSPAN, task-switching, timed typing, and an English interview for non-native speakers—while simultaneously recording ECG, temple PPG, and reference cuff BP. (The experiment was approved by the appropriate institutional ethics committee) Results and Future Work: Stress tasks induced systolic BP elevation in most participants (approximately 5–10 mmHg), with the English interview producing the largest response (mean increase:13.65 mmHg). Participants with higher resting BP showed greater stress-induced variability. Long-term monitoring was completed with acceptable comfort reported by participants. These results suggest that the temple-PPG system captures physiologically meaningful BP dynamics across different stress conditions, supporting its feasibility for long-term monitoring. Future work will focus on developing and validating cuffless BP estimation models using the collected ECG–PPG data and extending their application to long-term daily-life recordings.

Nr: 72
Title:

Towards Non-Destructive 3D Microstructural Analysis: A Comparison of Multi-Photon Microscopy and Contrast-Enhanced Micro-Computed Tomography

Authors:

Clarissa Silke Holzer Stock, Maïté Pétré, Anna Pukaluk, Isabelle Gennart, Nele Famaey, Greet Kerckhofs and Gerhard A. Holzapfel

Abstract: Cardiovascular diseases, the leading cause of death worldwide, are associated with critical conditions of the arterial walls. For an accurate diagnosis and effective treatment plan, a thorough characterization of the complex microstructure is crucial to decipher its (dys)functionality in health and disease and to relate the latter to disease progression. Currently, classical 2D histology is the gold standard for microstructural imaging in clinical practice and research. However, its destructive nature and thereon-based limited spatial information are problematic given the intricate architecture of highly heterogeneous biological samples, especially the arrangement of the extracellular matrix (ECM). In this work, we investigated the feasibility and utility of two non-destructive imaging techniques for 2D+ and 3D histology, multi-photon microscopy (MPM) and contrast-enhanced computed tomography (CECT), including its cryogenic version, for characterizing arterial architecture. We developed and optimized a sequential imaging and analysis protocol to image the same porcine and ovine arterial samples using both modalities. This approach enabled us to compare both imaging techniques and identify strengths and weaknesses of each in terms of visualization and quantification of ECM components, including, among others, elastin and collagen, in porcine and ovine arteries. MPM, particularly two-photon excitation fluorescence and second-harmonic generation signals, enabled high-resolution imaging of and clear differentiation between elastin and collagen fibers, allowing for the determination of their main characteristics, e.g., thickness, straightness, and orientation. (Cryo-)CECT, on the other hand, provided full 3D visualization and spatial quantification of the interlamellar spacing and volume fraction of adipocytes, medial elastic lamellae, and adventitial collagen fibers. Importantly, cryo-CECT was applied to arterial tissue for the first time, showcasing enhanced image sharpness and collagen visualization. After confirming the comparability of the two imaging techniques, the sequential imaging protocol enabled us to conduct a species- and vessel-specific comparison study. Notwithstanding the differences in resolution, sample preparation, and analysis approaches, both modalities showed consistent trends in vessel wall characteristics. Resembling human aortic tissue, the porcine samples showed a highly organized lamellar structure and an overall higher elastin and collagen volume fraction. Ovine aorta and pulmonary arteries, in contrast, featured medial gaps and disorganization of the lamellar structure, which casts doubt on the suitability of the sheep model as representative of the human vascular system in research. In summary, this study investigated the strengths and weaknesses of MPM and (cryo-)CECT and provided valuable insights into their respective contributions to the qualitative visualization and quantitative analysis of arterial tissue. Given their broad applicability in the field of biological soft tissues, this multimodal comparative approach should therefore serve as a first step toward advanced, non-destructive 3D histology and an improved understanding of soft tissue architecture in health and disease.

Nr: 84
Title:

Modeling the Enablement of Healthcare Big Data in the Context of E-Business Services

Authors:

Jolanta Kuklyte

Abstract: Healthcare systems generate vast volumes of data, yet much of it remains underutilized in supporting patient-centered care and preventive services. While e-business platforms such as telemedicine, patient portals, and digital health marketplaces depend on effective data integration, current approaches rarely provide a unified model linking healthcare big data with e-business services. This research addresses the problem of how healthcare big data can be effectively enabled and modeled within the context of e-business. A design science approach is applied, beginning with a review of existing frameworks in big data enablement, healthcare informatics, and e-business integration. The analysis reveals a lack of integrated models that simultaneously address technical, organizational, and service-delivery perspectives. A layered conceptual model includes healthcare data sources; integration and interoperability mechanisms; analytics and decision-support functions; e-business service delivery channels, and healthcare outcomes. This framework provides a structured pathway from raw data to actionable digital health services. The expected contribution lies in offering healthcare providers, policymakers, and technology developers a reference model that supports interoperability, safeguards privacy, and strengthens patient engagement. Future work will focus on validating the model empirically and exploring the role of emerging technologies, such as artificial intelligence, machine learning, and the Internet of Things, in enhancing healthcare big data enablement.

Nr: 124
Title:

Enhancing Fluorescence Microscopy Sensitivity Using Deuterated Water: Mechanisms and Applications

Authors:

Petr Táborský, Josef Kučera and Lenka Mádi

Abstract: Fluorescence microscopy is a cornerstone technique in chemical and biological research, offering high sensitivity and spatial resolution for the visualization of molecular and cellular structures. However, its performance is often limited by fluorescence quenching effects, particularly in aqueous environments. Water (H₂O), the most commonly used solvent in biological systems, is known to quench fluorescence through vibrational energy transfer and hydrogen bonding interactions. This study explores the substitution of H₂O with deuterated water (D₂O) as a strategy to enhance fluorescence signal intensity and detection sensitivity in fluorescence-based methods, with a particular focus on fluorescence microscopy and capillary electrophoresis with laser-induced fluorescence detection (CE-LIF). The enhancement of fluorescence in D₂O arises from several mechanisms. First, the isotopic exchange of labile hydrogens (e.g., –OH, –NH₂) with deuterons leads to reduced non-radiative decay pathways due to stiffer vibrational modes. Second, the lower vibrational energy of D₂O compared to H₂O minimizes dipole-dipole energy transfer from excited fluorophores to solvent molecules. Other types of mechanisms were also investigated in this work. These effects collectively result in increased fluorescence lifetimes and quantum yields, particularly for fluorophores emitting in the red region (>500 nm), which are commonly used in microscopy. Experimental data demonstrate that D₂O significantly enhances fluorescence intensity across a wide range of fluorophores, including DNA intercalating dyes (e.g., ethidium bromide, propidium iodide), fluorescent proteins (e.g., mCherry, GFP), and pharmaceutical compounds such as anthraquinones and anthracyclines. In fluorescence microscopy, D₂O improves image brightness, signal-to-noise ratio, and reduces photobleaching, enabling lower excitation power and better viability in live-cell imaging. Super-resolution techniques such as dSTORM and FLIM also benefit from increased photon yield and extended fluorescence lifetimes in D₂O-buffered samples. In CE-LIF applications, the use of D₂O as the background electrolyte solvent leads to markedly lower limits of detection (LOD) for analytes such as doxorubicin, rhein, and aloe-emodin—up to 40-fold improvement compared to H₂O. The enhanced sensitivity is attributed to both increased fluorescence intensity and reduced baseline noise in D₂O. Despite its higher cost, the minimal volume requirements in CE make D₂O a cost-effective choice for high-sensitivity analyses. Overall, the substitution of H₂O with D₂O presents a simple yet powerful approach to improve the performance of fluorescence microscopy and related techniques. The findings support broader adoption of D₂O in analytical workflows where fluorescence sensitivity is critical, offering enhanced detection capabilities without altering the chemical structure of the analytes or requiring complex instrumentation.

Nr: 351
Title:

Unlocking Predictive Insights from Neonatal Oxycardiorespirograms: Machine Learning for Respiratory Pattern Differentiation

Authors:

Kora Sabine Helm, Alexandra Epp, Michael Gerstlauer, Michael K. Baumgartner, Ludwig C. Hinske, Melanie L. Conrad, Mathias Kaspar and Fabian B. Fahlbusch

Abstract: Central apneas are a common and clinically relevant phenomenon in preterm neonates, reflecting immaturity of central respiratory regulation. Differentiating central from obstructive events and assessing therapeutic response (to caffeine treatment) requires valid, non-invasive monitoring strategies. Polysomnography, while comprehensive, is technically challenging in this fragile population. Multi-channel oxycardiorespirograms (OCRG) offer a feasible alternative, simultaneously capturing respiratory patterns, heart rate variability, and oxygen saturation in neonatal intermediate care. At the Department of Pediatrics and Adolescent Health, University Hospital Augsburg, Germany, OCRG recordings are routinely used to assess cardiorespiratory maturity and guide therapeutic decisions, including the timing of caffeine therapy discontinuation. Yet, despite integration into routine clinical care, standardized analytic frameworks for OCRG interpretation remain limited. Machine learning offers the opportunity to move beyond descriptive reporting, enabling the data-driven identification of respiratory patterns, artifacts, and treatment responses. We retrospectively analyzed 296 OCRG recordings (2021/01–2025/06), selecting one per patient to avoid follow-up bias, resulting in 75 recordings from neonates with a corrected age <52 weeks (17 preterm, 58 term; 8 vs. 9 and 27 vs. 31 female and male patients, respectively). In addition to gestational age, clinical parameters such as mode of delivery, antenatal corticosteroid therapy, caffeine treatment, respiratory support, surfactant administration, birth weight, and sex were recorded. After signal preprocessing and normalization (resampled to 5 Hz), we explored nasal flow channels as the primary diagnostic input and will extend the analysis to multimodal OCRG signals. Preliminary clustering of preprocessed data revealed distinct waveform patterns, supporting the feasibility of automated feature extraction. Building on this, we will evaluate supervised learning models for pattern classification and artifact detection. Explainable AI (XAI) methods, such as SHAP and Integrated Gradients, will be applied to ensure clinical interpretability. Evolving OCRG into a tool for clinical decision support, optimizing surveillance strategies, and contributing to safer discharge planning in vulnerable neonates.

Nr: 359
Title:

Detection of Population-Related Signals from Gut Microbiome Using Network-Aware Compositional Analysis

Authors:

Koji Ishiya

Abstract: Human gut microbiomes may inherently encode aspects of host populations and lifestyle environments, allowing the possibility of microbiome-based host attribute inference such as ethnic or regional classification. These communities reflect long-term dietary patterns, ecological exposures, hygiene practices, sociocultural structures, and potentially host genomic background, suggesting that they may retain population-specific signals beyond individual variation. However, taxonomic count data are compositional, sparse, and shaped by inter-taxon dependencies, making it difficult to extract stable and interpretable population-level patterns using conventional methods. Building on our previous network-based compositional framework proposed by Ishiya and Aburatani (2022, Physical Biology), I further extend this approach to enable population attribute inference from taxonomic abundance profiles. The workflow begins with zero adjustment through multiplicative replacement followed by centered log-ratio transformation, maintaining the principles of compositional data treatment used in the earlier method. Conditional dependency structures among taxa are then estimated to construct microbe–microbe interaction networks. Community detection applied to these networks yields microbial modules representing sets of co-varying genera with shared ecological or functional tendencies, expanding the earlier framework to incorporate population-discriminative structure. These module-level features are combined with transformed abundances to support multiclass population attribute inference. Complementary modeling strategies emphasize interpretability and reproducibility, enabling the identification of genera and modules consistently contributing to population distinctions. The extracted signatures align with known geographic, dietary, and cultural differences, and the module-based representation provides an interpretable view of coordinated microbial behavior not evident at the level of individual taxa. The workflow demonstrates robustness across datasets from distinct human populations, indicating that the signatures reflect stable attributes rather than cohort-specific noise. By extending the earlier network-aware compositional method to population-level inference, this framework provides a principled approach for characterizing human gut microbiome variation and highlights the potential of microbiome-based attribute analysis for studies of epidemiology, environmental exposure, migration, and sociocultural transitions.

Nr: 383
Title:

Trends and Key Patterns in the Prescription of Medicinal Products for the Treatment of Orphan Diseases among Statutorily Insured Patients in Schleswig-Holstein

Authors:

Reinhard Schuster, Timo Emcke, Marc Heidbreder and Mareike Burmester

Abstract: Orphan drugs are medicinal products intended for the treatment of rare diseases for which development under standard market conditions is challenging. They are developed for patients suffering from very severe conditions for which no, or at least no satisfactory, treatment options are currently available. Only a small proportion of the population—often affected since birth or early childhood—is impacted by these diseases (defined in Europe as a prevalence of fewer than 1 in 2,000 individuals), which corresponds to approximately 1,500 patients in Schleswig-Holstein. The data basis comprises prescription data from all statutorily insured individuals who received a medicinal prescription from a Statutory Health Insurance physician in Schleswig-Holstein. This includes approximately 1.5 million patients per year. The present analysis represents an update and extension of a previous evaluation covering a ten-year period (cf. conference abstract and poster presented at GMDS 2025 conference). There are considerable differences in the number of newly developed medicinal products in this field that have obtained market access (2015: 15; 2020: 87; 2024: 64). The highest market penetration during the first two quarters of 2025 was observed for orphan drugs approved in 2020, with revenues of EUR 10.14 million, followed by approvals in 2022 (EUR 2.21 million), 2024 (EUR 1.94 million), 2018 (EUR 1.54 million), and 2023 (EUR 1.02 million). Orphan drugs entering the market in 2022 were primarily assigned to ATC code L01FC01, accounting for EUR 10.0 million in revenue, followed by ATC code N07XX08 with EUR 4.4 million. The leading ATC code overall in 2022 was L01EX18, with revenues of EUR 1.28 million. ATC code L01FC01, with the active substance daratumumab, refers to the medicinal product Darzalex, which is used for the treatment of diseases involving the accumulation of misfolded proteins in the body (cost: EUR 5,809). Darzalex had already received EMA approval in 2016 for the treatment of multiple myeloma, a bone marrow malignancy characterized by the overproduction of malignant plasma cells. ATC code N07XX08, with the active substance tafamidis, refers to the medicinal product Vyndaqel, used for the treatment of transthyretin amyloidosis, either hereditary or wild-type, characterized by the accumulation of misfolded transthyretin in organs (cost: EUR 11,778). ATC code L01EX18, with the active substance avapritinib, refers to the medicinal product Ayvakit, used for the treatment of gastrointestinal stromal tumors, i.e. sarcomas of gastrointestinal tissue (cost: EUR 20,241). In summary, although the approval of new orphan drugs occurs in waves, an overall upward trend can be observed, accompanied by a broadening of the associated therapeutic spectrum. For future research, it would be of particular interest to investigate which pharmacological treatments patients received prior to the market introduction of an orphan drug that was therapeutically appropriate for them. Further relevant aspects include analyses of age structures stratified by active substance, as well as the medical specialties of prescribing physicians—especially with regard to whether prescriptions initiated by specialists can subsequently be continued by general practitioners. This is unlikely to apply to medicinal products with very small patient populations but is highly relevant for drug groups that are expanding into broader use.

Nr: 401
Title:

Intraoperative Gamma Probe–Guided Resection of High-Grade Gliomas Using 99mTc-MIBI: A Nuclear Medicine Bioimaging Approach

Authors:

Marcelo Mamede, Gabriel Filipe Soares Quiuqui, Lucas Rodrigues Souza, Bárbara Moreira Diniz, Antonio Augusto Carvalho Duarte, Caio Cesar Martins Pedrosa Castro, Mateus Pereira, Pablo Vinícius Nicoletti Ramos, Eduardo Paulino Júnior, Rodrigo Modesto Gadelha Gontijo and Andréa Lima Bastos

Abstract: High-grade gliomas are aggressive primary brain tumors associated with poor prognosis and limited overall survival. Surgical resection remains a cornerstone of treatment, and the extent of tumor removal is a critical determinant of clinical outcome. However, achieving maximal safe resection is particularly challenging due to the infiltrative growth pattern of these tumors and the limited ability of visual inspection and conventional neuronavigation to reliably distinguish tumor tissue from normal brain parenchyma during surgery. In this context, intraoperative bioimaging techniques that provide real-time molecular information may significantly enhance surgical guidance. This study evaluates the feasibility and diagnostic performance of intraoperative gamma probe detection following administration of technetium-99m methoxyisobutylisonitrile (99mTc-MIBI) as part of a multimodal, bioimaging-guided approach for high-grade glioma resection. A non-randomized clinical trial is ongoing at a public oncology hospital in Minas Gerais, Brazil. At this moment, thirty-two patients with suspected high-grade gliomas were screened, and twenty-six patients underwent craniotomy using standard microsurgical techniques. Among these, twenty-three patients received intravenous 99mTc-MIBI within a maximum interval of 10 hours prior to surgery and were evaluated intraoperatively using a handheld gamma detection probe. During the surgical procedure, radiation count rates (counts per second) were systematically acquired from tumor tissue and adjacent peritumoral brain regions. These intraoperative measurements were correlated with histopathological findings to assess the ability of gamma detection to differentiate tumor from non-tumoral tissue. In this interim analysis, intraoperative gamma probe guidance demonstrated a sensitivity of 92.9% for identifying tumor tissue. Specificity could not be reliably estimated at this stage due to the limited sample size and ongoing data collection. Notably, higher signal-to-background ratios were observed when the radiotracer was administered closer to the time of surgery, indicating a time-dependent optimization of detection performance. The results demonstrate that intraoperative gamma detection using 99mTc-MIBI is a feasible and sensitive nuclear medicine–based bioimaging tool for guiding high-grade glioma resection. The integration of molecular imaging information with real-time intraoperative detection has the potential to improve surgical precision and support multimodal image-guided neurosurgery. These preliminary findings support further investigation in larger cohorts to validate diagnostic performance, establish quantitative thresholds, and refine clinical workflows for routine implementation.

Nr: 410
Title:

Imaging-Guided Multimodal Fusion of microCT and Histology Using Contrast-Tunable 3D-Printed Reference Structures

Authors:

Philipp Nolte, Chris Johann Ackurat, Marcel Brettmacher, Marius Reichardt, Marieke Stammes, Christoph Rußmann and Christian Dullin

Abstract: Multimodal fusion of three-dimensional X-ray imaging and two-dimensional histology enables comprehensive structural and contextual analysis of biological specimens, but accurate registration of histological sections within volumetric micro-computed tomography (microCT) datasets remains a major technical challenge. Existing approaches often rely on manual alignment or computationally expensive intensity-based registration methods, limiting robustness and scalability. Reliable reference structures that are consistently visible across imaging modalities are therefore essential for simplifying and automating multimodal image fusion. In this work, we present an imaging-driven workflow that employs contrast-tunable, three-dimensional printed reference structures to enable precise and efficient alignment of microCT and histological data. Conic reference markers with adjustable X-ray attenuation were fabricated using digital light processing printing and embedded alongside biological tissue in resin blocks. Prior to sectioning, samples were imaged using propagation-based phase-contrast microCT, providing high-resolution volumetric context. Following automated sectioning and histological processing, whole slide images were acquired and integrated into the three-dimensional microCT coordinate system using the reference structures as fiducial landmarks. Image processing pipelines were developed to segment the markers in both modalities using classical image analysis methods as well as deep learning–based foundation models. The reference structures exhibited strong and tunable contrast in microCT volumes and remained clearly identifiable in histological sections, enabling robust segmentation and reliable landmark extraction. This allowed accurate, repeatable, and largely automated two-dimensional to three-dimensional registration of histological images, substantially reducing manual intervention and computational complexity compared to conventional approaches. The proposed methodology provides a practical and scalable solution for multimodal correlative imaging and establishes a robust framework for high-throughput integration of microCT and histology, with potential applications in computational pathology, three-dimensional histological reconstruction, and machine learning–based image registration.

Nr: 411
Title:

PBI-microCT–Guided Automated Punch Biopsy for FFPE-Embedded Specimens

Authors:

Philipp Nolte, Chris Johann Ackurat, Edina Ringleben-Fricke, Marius Reichardt and Christoph Rußmann

Abstract: In conventional CT-based diagnostics, spatial resolution is often insufficient to visualize small or deeply embedded biological structures, which may therefore remain undetected in standard scans. High-resolution analysis typically requires micro-computed tomography of biopsy punches extracted from paraffin-embedded specimens, yet current punch workflows rely on surface histology to define regions of interest, limiting accurate targeting of subsurface and three-dimensional structures. To overcome this limitation, we present an imaging-guided workflow that replaces surface-based targeting with volumetric propagation-based phase-contrast microCT. Regions of interest are selected directly within the three-dimensional microCT volume, enabling precise localization of internal tissue structures prior to extraction. The image-derived target coordinates are then translated into physical punch positions using an automated system built on a repurposed, non-functional consumer 3D printer, which serves as a low-cost, high-precision motion platform. This system enables automated and reproducible positioning of a punch biopsy tool, directly linking in silico microCT targeting with physical tissue extraction from embedded specimens. Initial precision tests using reference structures demonstrate accurate correspondence between image-defined targets and punch locations, while validation on biological specimens is ongoing. Overall, the presented approach enables precise and automated microCT-guided punch biopsies, reduces manual intervention and specimen loss, and establishes a scalable foundation for future autonomous tissue extraction workflows that tightly integrate volumetric imaging and histological analysis.

Nr: 415
Title:

Methodology for Development of Microfluidic Channels Integrated into Microfluidic Devices for Blood Cells’ Investigations

Authors:

Tihomir Tiankov, Anika Alexandrova-Watanabe, Emilia Abadjieva, Dimitar Trifonov, Yavor Tsokov, Assen Shulev and Svetla Todinova

Abstract: 3D printing of microfluidic devices provides design freedom, which is difficult or impossible to achieve using traditional manufacturing techniques. This allows the development of truly three-dimensional microfluidic designs and the fabrication of intricate and customizable vascular networks that can mimic the complexity of natural blood vessels The work aims to present 3D modelling, simulation and development of 3D nanoprinted microfluidic channels. A methodology, combining two main technologies as 3D nanoprinting and soft lithography, has been developed for suitable prototyping of microfluidic channels that could be integrated in a microfluidic device for blood cells’ investigations. The 3D design of a microfluidic channel is realized by the 3D Computational Aided Design (CAD) analysis software - SOLIDWORKS. A suitable laminar flow is generated using Computational Fluid Dynamics (CFD) software. Flow 3D commercial software platform and OpenFOAМ free software platform are used for simulation. The obtained results, including pressure, velocity, and wall shear stress in the microfluidic channel, have been compared and analyzed. By Flow 3D platform also simulated the flow of aggregation of red blood cells (RBCs) in the developed microchannel. RBCs are modeled as solid particles, and the size and shape of the RBC aggregates specified.The fluid flow is simulated, setting the same density of the suspension used for real experiments. The collisions between the spherical particles are investigated. The contact between the particles and the walls of the micro-channels is also simulated. The real physical prototype of a microfluidic channel is developed using 3D nanoprinting by two-photon polymerization technology by Photonic Professional GT2 (Nanoscribe, Germany) equipment and soft lithography station for the development of microfluidic devices from photoresist materials (Elveflow, France). 3D nanoprinting is used for the development of molds from photosensitive resins used for soft lithography micro-molding. Two types of photosensitive resins are used during 3D nanoprinting. Different supports as silicon wafers and cover glasses, are tested for better adhesion of the developed polymer matrices. The polydimethylsiloxane (PDMS) is poured onto the mold, and after backing, an elastic casting is obtained. Using confocal and optical microscopy, the shape of the resulting microchannels is investigated for optimal adjustments. Experiments testing the functionality of the microfluidic channels are performed with dilute suspensions of RBCs. Different pressures are applied, and the flow RBC suspension in which aggregates are formed is analyzed. The results obtained during the experiments are compared with the results of the simulation. The present study offers an important technological approach to the rapid production of low-cost microfluidic channels, with the help of which the rheological properties of RBCs can be studied in various diseases. Acknowledgements: This work is realized in the Center of Competence for Mechatronics and Clean Technologies “Mechatronics, Innovation, Robotics, Automation and Clean Technologies” – MIRACle, with the financial support of contract No. BG16RFPR002-1.014-0019-C01, funded by the European Regional Development Fund (ERDF) through the Programme “Research, Innovation and Digitalisation for Smart Transformation” (PRIDST) 2021–2027.

Nr: 421
Title:

Unified Software Application for Automated Image Flow Analysis of Red Blood Cell Aggregates in Pathological Conditions: A Color-Coded Classification Approach

Authors:

Emilia Abadjieva, Anika Alexandrova-Watanabe, Tihomir Tiankov, Ariana Langari, Miroslava Ivanova, Dimitar Trifonov and Svetla Todinova

Abstract: Quantitative flow image analysis of RBC aggregation is often time-consuming and subjective, posing significant challenges in biomedical research. Conventional manual evaluation of RBC rheology is imprecise and inconsistent, particularly in pathological conditions such as preeclampsia and leukemia. To address this limitation, an ImageJ-based automated image flow analysis software is developed for faster, more accurate, and efficient determination of RBC rheological characteristics through color-coded classification of RBC aggregates. The software system is based on binary image processing combined with particle analysis to detect and measure RBC aggregates ranging from 50 to over 4000 µm². The algorithm incorporates steps including gap closure and hole filling for accurate object segmentation. Aggregates are classified into six size categories, with each category assigned a distinct color: red (50-150 µm²), green (150-500 µm²), blue (500-1000 µm²), cyan (1000-2000 µm²), magenta (2000-4000 µm²), and yellow (>4000 µm²). Each aggregate is marked with its corresponding color while the software counts aggregates in each category and calculates their total area, generating data tables. This classification enables differentiation between normal rouleaux formations (50-500 µm²), medium 3D clusters (500-2000 µm²), and large complex aggregate networks (>2000 µm²). The analysis of RBC aggregates in the considered intervals reveals distinct pathological changes. The software is applied to analyze RBC aggregation in different pathologies – preeclampsia (PE) and chronic lymphocytic leukemia (CLL). In the PE studies, significantly increased RBC aggregation is observed. The aggregation index (AIRBC) for the non-severe PE increases 26% compared to the AIRBC, while for the severe PE, AIRBC rises 75%, and characterized with smaller number of disaggregates at high shear rates, indicating formation of stable RBC clusters. In CLL studies, untreated patients show considerably increased RBC aggregation with large 3D network structures not observed in healthy controls. The target treatments of CLL with Obinutuzumab/Venetoclax shows no improvement, whereas Ibrutinib successfully restores aggregation levels comparable to healthy individuals. The researches of disaggregation illustrate a critical shear rate of 446 s⁻¹ for complete dispersion. In CLL patient samples, a significant reduction in RBC aggregation at high shear rates are observed, with an increased number of stable aggregates, indicating an impaired RBC disaggregation capacity. This color-coded software image flow analysis provides a reliable tool for quantitative assessment of RBC aggregation. This software can be applied to the quantitative assessment of RBC aggregates not only for the pathologies PE and CLL, studied by us, but also in other diseases, in which changes in the blood rheology are observed. Acknowledgement: This work is realized in the Center of Competence for Mechatronics and Clean Technologies “Mechatronics, Innovation, Robotics, Automation and Clean Technologies” – MIRACle, with the financial support of contract No. BG16RFPR002-1.014-0019-C01, funded by the European Regional Development Fund (ERDF) through the Programme “Research, Innovation and Digitalisation for Smart Transformation” (PRIDST) 2021–2027. This work also is supported by Grant KP-06-H73/3, competition for financial support for basic research projects—2023, Bulgarian National Science Fund.

Nr: 424
Title:

Neuro-Astrocytic Computational Framework for Somatosensory Dysfunctions in Eating Disorders: Integrating Biophysical Models, Virtual Reality, and Machine Learning

Authors:

Gaia Roccaforte, Rawan Mahmood Ahmad Nawaiseh, Rosa Musotto, Giovanni Pioggia and Dario Bruneo

Abstract: Eating disorders (EDs), particularly anorexia nervosa, are characterized by profound disturbances in body perception and by neurobiological alterations involving both neuronal and glial components of cortical circuits. Evidence from experimental models suggests that reductions in astrocyte density and impaired astrocyte-neuron interactions may contribute to synaptic dysregulation and circuit instability, potentially underlying abnormal responses to body and food-related stimuli. This work introduces a preliminary, conceptually integrated neuro-informatic framework that combines a biophysical neuron-astrocyte model, a Virtual Reality (VR) based phenotyping platform equipped with wearable biosensors, and a machine learning pipeline to investigate ED-related circuit dysfunctions. At the computational level, we employ an extended Hodgkin-Huxley type model incorporating astrocytic modulation to simulate how decreased astrocyte density and reduced synaptic regulation efficiency may destabilize cortical microcircuits, leading to hypo- or hyper-reactive network dynamics. At the behavioral and physiological level, a VR platform composed of two modules, Body Swap and Food Exposure, was developed using the Unity Engine and integrated with textile-based wearable biosensors to quantify embodiment-related behavior and physiological responses linked to autonomic regulation. In this preliminary phase, the platform was evaluated in a usability study involving twelve healthy female participants, demonstrating good usability and acceptability. Finally, using synthetic data and model-based simulations, we outline how multimodal features derived from computational dynamics, VR behavioral measures, and physiological signals could be integrated within a machine learning framework to classify ED-like profiles and explore neurobehavioral biotypes. These results are presented as proof-of-concept demonstrations rather than empirical clinical findings. Overall, this multidisciplinary framework bridges bioengineering, neuroscience, and clinical psychology, positioning VR as a potential computational phenotyping tool capable of linking cellular-level dysfunctions to measurable behavioral and physiological biomarkers, with implications for the future development of personalized digital therapeutics for eating disorders.

Nr: 426
Title:

3D Design, Modeling and Prototyping of PDMS Microfluidic Devices for Blood Cells Investigation

Authors:

Dimitar Trifonov, Tihomir Tiankov, Anika Alexandrova-Watanabe, Emilia Abadjieva, Yavor Tsokov, Pavel Venev, Assen Shulev, Aleksandar Iliev and Svetla Todinova

Abstract: Inertial microfluidic technology is innovative and has been successfully used in separation processes when working with biological cells, such as the separation of sperm cells from samples containing high concentrations of white and red blood cells (WBC and RBC). Various microfluidic studies are developed in the last decade, focused on the deformation of blood cells flowing through micro-channels mimicking micro-vessels in-vivo. The deformability of the RBC in dynamic flow could be investigated successfully in specially developed microfluidic devices. The proposed methodology has significant advantages over other currently known technologies in terms of reducing the time for the process, no externally applied forces, low energy consumption, low cost and short time for the preparation of microfluidic devices. Recently, experiments with such devices using a spiral-shaped microfluidic channel have been reported in the literature. The aim of this study is to present 3D modeling and simulation of inertial PDMS microfluidic devices with a serpentine-shaped micro-channel. Various manufacturing methods of the mold are discussed, analysed and applied for development of micro-channels with different shape and size suitable for integration in microfluidic devices for blood cells investigations. 3D models of a serpentine microfluidic channel with different channel sizes were developed using a Computer-Aided Design (CAD) software platform SolidWorks. The 3D models were simulated in Computational Fluid Dynamics (CFD) software FLOW3D, investigating the behavior of the microfluidic flow at different pressures. The development of a suitable method for creating molds in a spiral shape for casting a polymer material as PDMS consists of design of the spiral micro-channel in different sizes. The development of a spiral with a channel width, more than 0.5 mm uses a standard 3D printer (Anycubic Photon Mono X). In the development of a spiral with a channel width of about 0.2 mm and less, a femptosecond laser Micro-machining Workstation FemtoLAB (Lithuania) is used. The mold used to develop the spiral with bigger size of the micro-channel consists of a polymer frame, a polymer substrate with a 3D printed spiral shape, and a glass top cover. The mold used for developmnt of the spiral with the narrower micro-channel consists of a polymer 3D printed frame that is closed at the top and bottom with two glass plates. After the creation of the PDMS plate with parallel walls, a spiral shape is developed on it using the femtosecond laser workstation. Two methods are used to pour the PDMS material into the specified molds, in an open mold by a standard casting or by PDMS micro-molding. As a result, a detailed technical methodology is created for the development of microfluidic devices of various shapes and sizes from PDMS in order to reduce the time and cost of production. Prototypes of microfluidic channels are developed finally following the methodology proposed. Acknowledgements: This work is realized in the Center of Competence for Mechatronics and Clean Technologies “Mechatronics, Innovation, Robotics, Automation and Clean Technologies” – MIRACle, with the financial support of contract No. BG16RFPR002-1.014-0019-C01, funded by the European Regional Development Fund (ERDF) through the Programme “Research, Innovation and Digitalisation for Smart Transformation” (PRIDST) 2021–2027.

Nr: 427
Title:

Physics-Informed Statistical Quantification for Magnetic Resonance Spectroscopy

Authors:

Ryo Enari, Hiroyuki Ueda, Yosuke Ito and Kazuyoshi Yoshii

Abstract: Magnetic resonance spectroscopy (MRS) enables a non-invasive measurement of metabolite concentrations from spectral signals. Despite its potential, reliable quantification across different acquisition conditions remains a fundamental challenge. Variations in magnetic field inhomogeneity, sequence parameters, and hardware settings often lead to inconsistent estimates, limiting reproducibility in multi-site and longitudinal studies. This limitation originates from the conventional formulation of MRS quantification as a signal approximation problem. Standard approaches, such as linear combination modeling, rely on simplified assumptions of the signal generation process and treat physical parameters as fitting coefficients. As a result, the inverse estimation becomes implicitly dependent on observation conditions, even when the underlying physical or biological state is unchanged. The proposed framework is formulated as a physics-informed autoencoder tailored for MRS signal modeling. The encoder takes time-domain MRS signals as input and estimates a set of physically interpretable spectral parameters using a one-dimensional convolutional neural network (1D-CNN). These parameters include metabolite concentrations, relaxation times, frequency offsets, and phase terms, which fully characterize the underlying signal generation process. Instead of employing a trainable neural decoder, the estimated parameters are passed to a fixed physics-based decoder. This decoder reconstructs the MRS signal by explicitly solving the Liouville–von Neumann (LVN) equations, generating a spectral basis in the time domain that is subsequently transformed into the frequency domain. The fidelity of signal reconstruction is evaluated using loss functions defined in both the time and frequency domains, enforcing consistency across complementary representations of the same physical process. This design enables self-supervised learning without requiring paired ground-truth. Because the reconstruction target is the observed signal itself and the decoder strictly follows the forward physical model, learning is guided by physical consistency rather than statistical similarity. As a result, the encoder is trained to infer intrinsic physical parameters that are invariant to acquisition-dependent variations. Crucially, this physics-informed framework decouples parameter estimation from site-specific observation conditions. Since differences across scanners, protocols, and field inhomogeneities are absorbed by the forward model, the encoder learns representations that generalize across acquisition environments. This property makes the proposed framework particularly suitable for multi-site MRS studies, where robustness and standardization are essential. By embedding the forward physical model directly into the autoencoder structure, this framework offers a structured alternative to purely data-driven spectral fitting. The proposed approach provides a scalable pathway toward robust and site-invariant MRS quantification.

Nr: 428
Title:

Cardiac Hypertrophy and Systolic Dysfunction in Mice with Smooth Muscle Cell-Specific Deletion of Fibulin-4

Authors:

Jungsil Kim, Woori Cho and Hee-Young Yang

Abstract: Fibulin-4 (Fbln4) is a critical glycoprotein essential for the proper assembly of elastic fibers in the extracellular matrix. The targeted deletion of Fbln4 specifically within smooth muscle cells (Fbln4SMKO) is known to induce severe ascending aortic aneurysms, which are characterized by 1.5 times greater outer diameter compared to the normal. The aneurysmal aorta in Fbln4SMKO showed the increased arterial stiffness [1], which subsequently imposes a chronic hemodynamic load on heart, impacting cardiac function and structural remodeling. In this study we hypothesized that the localized Fbln4 deficiency in the vasculature leads to progressive cardiac function impairment alongside significant structural changes. To test this, we utilized 3-month-old male mice (n=7~9) and compared them with wild-type (WT) littermates. All animals were anesthetized with isoflurane for non-invasive cardiac assessment. Cardiac morphology and in vivo function were evaluated using high-resolution transthoracic echocardiography (Vevo 2100, FUJIFILM VisualSonics Inc., USA). Contractile systolic parameters, including ejection fraction (EF) and fractional shortening (FS), were calculated from M-mode images obtained at the parasternal short-axis view. Additionally, diastolic function was meticulously assessed via tissue Doppler imaging, focused on the e' velocity and the E/e' ratio. Following imaging, the heart weight-to-body weight (HW/BW) ratio was measured to quantify hypertrophy. Histological sections were stained with Hematoxylin & Eosin (H&E) and Picrosirius Red to determine cardiomyocyte cross-sectional area and evaluate interstitial collagen deposition. Data was analyzed using an unpaired Student’s t-test, with statistical significance defined as p<0.05. Our results showed that mice developed marked cardiac hypertrophy, as indicated by significantly elevated HW/BW and left ventricular (LV) mass/BW ratios, as well as increased cardiomyocyte size in H&E stained images compared to WT controls. Functional analysis revealed a significant decline in systolic performance, with mice exhibiting substantially lower EF and FS values. While histological analysis showed only a mild increase in perivascular collagen, the biochemical assay confirmed a statistically significant increase in total myocardial collagen content in the hearts. These findings suggest that cardiac remodeling in mice is characterized by compensatory hypertrophy and systolic dysfunction. To further elucidate the mechanisms of this remodeling, the integrity of the aortic valve, which connects the aorta to the heart, warrants investigation. Given that heart valves are rich in elastic fibers, the extracellular matrix defects caused by Fbln4 deficiency may extend to the valvular structures, potentially leading to valvular regurgitation. Future studies will focus on characterizing these valvular changes and their contribution to the observed cardiac dysfunction, providing a more comprehensive understanding of the cardio-vascular linkage in Fbln4-deficient models. Acknowledgement: This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (2021R1I1A3059740)

Area 1 - DEMS

Nr: 433
Title:

Objective Wearable Monitoring Identifies Activity and Health Risk Patterns in Young Adults: Implications for Monitoring System Design

Authors:

Dabin Choi and Catherine Park

Abstract: Wearable-based monitoring systems are increasingly applied in preventive healthcare to support early identification of health risks in both clinical and general populations. Young adults are typically regarded as a low-risk group due to preserved physical function and low disease prevalence and are therefore less frequently prioritized for intensive health monitoring. As a result, health assessment in this population often relies on self-reported measures. However, such approaches may not adequately capture underlying behavioral patterns, including reduced physical activity or associated psychological vulnerability. From a monitoring system design perspective, examining discrepancies between objective sensor-derived data and subjective health perception is important for evaluating current monitoring approaches and informing the development of user-centered healthcare monitoring systems. As a wearable-based health monitoring case study, 26 healthy adults aged ≥19 years (mean age ± SD: 43.15 ± 18.9 years; 9 males, 17 females) were recruited from Yonsei University Mirae Campus, South Korea, and stratified into two age groups (<40 years, n = 11; ≥40 years, n = 15). Objective physical activity was monitored for two consecutive days using a triaxial accelerometer-based wearable sensor (ActiGraph LEAP) worn on the non-dominant wrist. Outcome measures included step count, activity-related energy expenditure, moderate-to-vigorous physical activity (MVPA), sedentary time, and light-intensity activity. Subjective health perception and mental health status were assessed using validated questionnaires: the EuroQol five-dimension five-level questionnaire (EQ-5D-5L), the EuroQol visual analogue scale (EQ-VAS), DSM-5–based screening tools, the Short Geriatric Depression Scale (SGDS), and the Mini Nutritional Assessment (MNA). Between-group comparisons were performed using nonparametric statistical tests. Objective wearable-derived measures demonstrated significant age-related differences in physical activity patterns. Participants aged ≥40 years exhibited 39% higher daily step counts and 60.39% greater moderate-to-vigorous physical activity compared with younger adults, despite greater sedentary time. Younger adults showed slightly higher light-intensity activity. However, activity-related energy expenditure was only marginally higher (+4.5%), indicating lower overall physical activity levels. Subjective health perception, assessed using the EQ-VAS, was significantly lower in younger adults (78.18 vs. 91.53, p = 0.0034), whereas EQ-5D index scores did not differ significantly between groups. Elevated risks of depression, insomnia, and nutritional deficiency were observed predominantly in the younger group and were consistent with objectively measured inactivity patterns. This study indicates that wearable-based monitoring systems can identify physical activity patterns and associated health risks missed by self-reports, particularly in populations traditionally considered low-risk. The observed age-group differences should be interpreted in the context of monitoring system design rather than as evidence of age-related health advantage. From a design and evaluation perspective, integrating objective sensor-derived data with subjective assessments may support more accurate health risk identification and inform the development of scalable, user-centered healthcare monitoring systems aimed at early detection and health promotion in young adult populations.

Nr: 434
Title:

Designing a Mobile Snus Cessation Application for Young Swedish Women Using Nudge Theory and Conversational AI

Authors:

Eunji Lee and Maja Vikla

Abstract: The trend of increased snus use among young females of Swedish origin has become a pressing public health issue. Moreover, the recent influx of those using tobacco-free snus pouches has added to this serious health problem. Among young Swedish females between the ages of 16-29 years, there was a 500% increase in the prevalence of snus use over a six-year period of 2018-2024. This is a serious health crisis. Online health solutions such as mobile health apps provide new avenues. This study examines the design and development of a mobile application for snus cessation that is specifically designed for young Swedish women in the 16- to 29-year-old age group. The application integrates the concepts of nudging with the benefits that could be derived from a chatbot that uses AI to facilitate conversations. Literature review, state-of-the-art of applications incorporating digital nudges and chatbots, and a semi-structured interviews with 5 young Swedish female snus users aged 18-28 were conducted for this study. With the key findings from the aforementioned three methods, a new mobile application has been designed and developed by combining digital nudging elements, such as milestone functionality and visualizing health progress, and an empathetic conversational interface offered by a tailored chatbot. User tests reveal that the five test subjects were highly positive towards the conversational AI offering, finding that it is encouraging. A majority of the participants also felt that the nudge elements, such as the health timeline and achievements, were encouraging enough to use the new application as an integral part of their routine. This study makes an important contribution to the area of digital health by showing that the potential exists within a combination of nudge theory and conversational AI to help snus users quit. Its results show that tailored, theory-based, AI-integrated mobile health interventions can have an important part to play in dealing with new nicotine habits among young Swedish women.

Area 2 - Scale-IT-up

Nr: 45
Title:

MARI: Co-Produced Digital Alcohol Intervention Tailored for Women Who Are Trying to Conceive, Pregnant, or Parenting

Authors:

Abi Rose

Abstract: Alcohol consumption during preconception, pregnancy, and early parenting stages poses significant health risks for women and their children. Yet this is a hidden health issue, attached to significant stigma, and existing interventions often fail to meet the nuanced needs of women in these life stages. This gap underscores the need for digital health solutions that are not only evidence-based but also responsive to the unique experiences and challenges faced by women in these stages. The MARI (Maternal Alcohol Resources & Information) intervention aims to address this need by offering a co-produced, scalable digital alcohol intervention tailored for women who are trying to conceive (TTC), pregnant, or parenting. MARI combines established behavioural change techniques (BCTs) with engaging, user-centred wellbeing tools and personal stories from women. Our coproduction approach is grounded in the lived experiences of women from a variety of backgrounds, ensuring the intervention resonates with and is accessible to a wide range of users. By incorporating diverse therapeutic components, MARI aims to increase engagement, trust, and adherence. MARI integrates digital tools for self-monitoring and goal-setting, as well as promoting awareness and reflection, thus empowering women to make informed decisions about their health and alcohol consumption. This intervention employs an intersectional design framework to address the complexities of gender, health, and social factors that influence alcohol consumption patterns in these key stages of women’s lives. Through a focus on co-design, the intervention is able to ensure that the content is not only evidence-based but also relevant to the lived realities of women in varying contexts. The inclusion of personal narratives from women helps bridge the gap between scientific evidence and personal experience, fostering a sense of community and shared understanding among users. MARI’s coproduction process and user-centred design also align with best practices in the development of scalable digital health tools. By emphasizing flexibility in content delivery and accessibility, MARI can be adapted to different settings (e.g. online, healthcare, prison/probation) making it a potentially powerful tool for addressing alcohol-related harm among women across diverse populations. Furthermore, the inclusion of real-world data on user engagement and outcomes can inform ongoing improvements to the intervention, ensuring that it evolves to meet emerging needs. MARI's potential for scalability and impact is grounded in its combination of well-established BCTs, culturally relevant content, and a flexible digital delivery model. As digital health tools continue to grow in popularity, MARI represents an innovative step forward in integrating behavioural change, user experience, and health equity in the design of women’s health interventions. This abstract will highlight key lessons learned from MARI’s coproduction process, discuss its alignment with the principles of intersectional design, and explore how the intervention can be scaled to diverse populations. The presentation will provide insights into how digital interventions can be co-designed and scaled to meet the diverse needs of women, offering valuable takeaways for researchers, practitioners, and policymakers looking to advance digital women’s health. MARI is funded by the National Institute of Health Research (NIHR207252)

Area 3 - uOrganChip

Nr: 422
Title:

3D-Printed Vessel-on-a-Chip System with Integrated Sensors to Investigate Endothelial Dysfunction

Authors:

David Wörle, Xenia Kraus, Christoph Westerhausen and Janina Bahnemann

Abstract: Cardiovascular diseases, largely driven by atherosclerosis, remain the leading cause of death worldwide. A central role in the development of atherosclerosis is played by factors such as impaired vascular tone regulation, increased permeability, and enhanced adhesion to the endothelium, which are collectively referred to as endothelial dysfunction. This dysfunction is influenced by multiple, still insufficiently understood risk factors. This study introduces a 3D-printed microfluidic vessel-on-a-chip model of the arterial wall designed to investigate endothelial dysfunction under physiologically relevant conditions. In the system, endothelial cells are cultured under flow on a semi-permeable membrane, enabling co-culture with smooth muscle cells, while integrated sensors allow real-time monitoring of key functional parameters. Electrodes permit assessment of barrier integrity by impedance spectroscopy, and an optical pH sensor provides information on inflammatory changes associated with endothelial dysfunction. Endothelial cells exposed to a shear stress of 6 dyn/cm² in the device exhibit cytoskeletal alignment and oriented cell–cell contacts in the direction of flow, mimicking in vivo conditions. Dynamically cultured cells show an increased barrier function compared to statically cultured cells, reflected by elevated impedance values. Furthermore, an inflammatory response triggered by the cytokine TNF-α results in a decrease in barrier function and pH, with larger changes after static cultivation of the cells. This highlights the need for physiologically relevant models and demonstrates the potential of the presented vessel-on-a-chip platform as a versatile tool to study mechanisms of endothelial dysfunction and vascular pathology.