Conference Agenda

Managing the pressures at the socket-residual limb interface represents one of the primary challenges in the design of transfemoral prostheses. Indeed, the physiological volumetric fluctuations of the tissues throughout the day induces an increase of them, resulting in patient discomfort. This work illus-trates the development and analysis of a TPU liner, starting from the investi-gation of a modified re-entrant meta-structure in order to improve the pres-sure distribution in the critical region. With this aim, finite element simula-tions were carried out to study the mechanical behavior of the liner under specific loading conditions. Results highlighted how the proposed configura-tion ensures a pseudo-constant pressure variation over the physiological range of limb volume changes. Overall, the developed solution represents an effective approach to enhancing the biomechanical performance of the inter-face component, with potential positive implications for the user’s quality life.

Skin cancer, especially melanoma, is a crucial public health issue where ear-ly diagnosis could drastically improve survival rates. This paper presents a comparative analysis of two lightweight convolutional neural networks, Effi-cientNetB0 and EfficientNetB3, optimized for mobile applications, to support primary dermatological screening. To address the class imbalance in the HAM10000 dataset, a cost-sensitive learning approach combined with dy-namic data augmentation was adopted, prioritizing a conservative balancing strategy. The models were then evaluated using a recalibrated decision threshold to maximize clinical sensitivity for malignant lesions. Analysis re-sults showed that the EfficientNetB3 architecture, combined with the downsampling adopted strategy, achieves better performance. The findings demonstrate the viability of deploying optimized, balanced AI models on computationally restricted devices for accessible triage.

Arterial bifurcations are common sites of cerebral aneurysm development, largely because of the complex hemodynamic environment generated by flow division. However, the relationship between bifurcation geometry and aneurysm formation, growth, or rupture remains incompletely defined. This study aimed to evaluate the association between quantitative aneurysm morphology and bifurcation geometry, which was defined as the combined spatial arrangement of branch angles and vessel diameters.

A total of 197 bifurcation cerebral aneurysms were analyzed and for each case, three-dimensional vascular reconstructions were used to extract aneurysm morphological parameters and bifurcation-related geometric features. Branch angles and vessel diameters were measured using a systematic centerline-based approach. A regression-based machine learning framework was then applied to test different predictive models and to assess the relationship between bifurcation geometry and aneurysm morphology. Ridge regression was selected as the final model because it provided the most stable balance between predictive performance and interpretability.

Among the models tested for the different aneurysm morphological variables, the model predicting neck diameter achieved the best predictive performance, as measured by the coefficient of determination. Daughter-vessel diameters were the strongest predictors, whereas angular variables alone showed only limited predictive value. In the separate feature-set analysis, diameter-based models explained substantially more variance than angle-based models. However, the combined angle-diameter model provided the most informative representa-tion of the local bifurcation configuration. These findings suggest that bifurcation geometry should be interpreted as an integrated angle-diameter configuration rather than as an isolated angular feature.

Abdominal aortic diseases are commonly assessed through Computed Tomogra-phy (CT) and Computed Tomography Angiography (CTA), which provide de-tailed anatomical information on aneurysmal dilatation, stenosis, thrombus for-mation, and aorto-iliac involvement. However, standard image-based evaluation remains mainly morphological and does not directly describe local blood-flow stagnation, recirculation, or altered flow organization. In this work, a patient-specific Computational Fluid Dynamics (CFD) workflow is proposed to com-plement Digital Imaging and Communications in Medicine (DICOM)-based ana-tomical assessment with functional hemodynamic information. Three abdominal aortic models were analyzed: a healthy reference aorta, a stenotic/thrombotic con-figuration, and an Abdominal Aortic Aneurysm (AAA). The pathological geome-tries were reconstructed from contrast-enhanced arterial-phase CT images and simulated in Simcenter STAR-CCM+ using the same inlet waveform and three-element Windkessel outlet parameters. After verification of the pressure response in the healthy model, a new post-processing approach, named Functional Stagna-tion Mapping (FSM), was applied to the pathological cases to visualize low-velocity core regions while excluding the near-wall low-speed layer. The results are presented through three-dimensional FSM maps, two-dimensional slices at critical sections, and streamline analysis. The proposed approach highlights slow-flow and recirculating regions that are not directly evident from DICOM mor-phology alone, providing an intuitive CFD-derived functional layer for the inter-pretation of complex abdominal aortic pathologies.

Parkinson’s disease (PD) is a neurodegenerative disorder characterized by motor impair-ments such as bradykinesia, rigidity and reduced movement coordination. Clinical evalua-tion is commonly performed through observational scales, including the Movement Disorder Society Unified Parkinson’s Disease Rating Scale (MDS-UPDRS), whose assessment remains partially subjective and clinician-dependent. This research work proposed a prelim-inary Vicon–OpenSim workflow for the quantitative biomechanical assessment of Parkinsonian motor tasks. A healthy subject was acquired using a marker-based Vicon motion capture system while performing repeated seated leg-raising movements inspired by the MDS-UPDRS Leg Agility task. The acquired trajectories were processed and transferred into an OpenSim musculoskeletal model to reconstruct lower-limb kinematics within a multi-body biomechanical environment. The proposed framework aimed to extract objective kinematic descriptors, including joint range of motion, angular values and symmetry indices, integrating optical motion capture systems with musculoskeletal models. It provides an additional and reproducible biomechanical framework for transforming clinically standard-ized motor tasks into objective joint-level descriptors, useful in future biomechanical inves-tigations involving patients with Parkinson’s disease.

Osseointegrated transfemoral prostheses directly anchor the implant into the residual femur, eliminating the complications of socket-based systems. How-ever, the mechanical behaviour of the bone-implant interface critically gov-erns periprosthetic bone remodelling. Most finite element studies simplify this interface as fully bonded, overlooking the spatially heterogeneous nature of os-seointegration. This study evaluates the influence of contact condition, fric-tional versus bonded, and implant geometry on periprosthetic bone remodel-ing, comparing three transfemoral stem designs: straight, planar-curvature, and double planar-curvature. An iterative finite element framework coupled with a gradient-enhanced bone remodelling algorithm was employed, using strain en-ergy density as the mechanical stimulus. Under bonded contact, all designs showed approximately 19% bone density reduction across most Gruen zones, with no meaningful differences among geometries. Frictional contact produced lower resorption (≈16% in Gz6 for BPC implant) and revealed clear geometric sensitivity: the patient-specific designs achieved less resorption (16–18% ranges,), while the straight one showed the widest density reduction (18–19%). These findings demonstrate that interface modelling assumptions critically af-fect simulation outcomes, and that patient-specific curved stem geometries may offer biomechanical advantages over straight designs.