Quinto seminario – Presentazione degli avanzamenti e dei risultati dei Progetti Pilota del Progetto DARE

PROGRAMMA DELL’EVENTO

11:30-12:00: Spoke3, WP2, T2.1 – A Digital Twin technology to monitor the risk of fragility bone fractures in osteoporotic patients (relatrice: Julia Szyszko)

Abstract: BBCT-hip is an in silico methodology that predicts the absolute risk of hip fracture at the time of CT scan (ARF0) using calibrated femoral CT scans together with a subject’s height and weight. The method combines a stochastic model that generates one million possible sideways fall impact forces with a finite element (FE) model that estimates femoral failure load based on principal strain criteria across 28 impact configurations. ARF0 is calculated as the ratio of impact forces causing fracture to the total number of simulated impact forces. This pilot study aims to evaluate BBCT-hip in a prospective clinical cohort of osteoporotic and osteopenic postmenopausal women, assessing its fracture-risk predictive performance against the current gold-standard – areal bone mineral density (aBMD).

12:00-12:30: Spoke3, WP2, T2.4 – Predicting the risk of bone fracture in patients with metastatic carcinoma (relatrice: Elisa Bruatto)

Abstract: This pilot study aims to develop a system for estimating the risk of femoral fracture in patients affected by bone metastases, which can serve as a clinical decision-support tool for treatment selection, with a specificity higher than the current gold standard, represented by Mirel’s score. The methodology involves the creation of patient-specific finite element models. Fracture simulations are performed in Ansys with an approach that enables crack propagation tracking through the progressive deactivation of elements that exceed a given strain threshold. Biomechanical parameters extracted from the simulations are used to stratify patients into high- or low-risk categories for fracture. With the aim of automating the pipeline for clinical application, a deep learning-based model has also been developed for automatic femur segmentation.

12:30-13:00: Spoke2, WP4, T4.3 – A sustainable and technological approach to large-scale prevention of falls and injuries (relatore: Giacomo Savelli)

Abstract: BoneStrength provides an in silico framework to estimate hip fracture incidence by integrating subject-specific and population-based descriptions of bone strength, fall exposure, and impact conditions. Its computational nature allows preventive strategies to be investigated before implementation in real-world settings, contributing to a sustainable approach in which fall-related injuries can be predicted at population scale. This approach supports the transition from purely observational risk assessment to predictive, scalable, and technology-driven prevention of fall-related injuries, particularly in older and osteoporotic populations.

13:00-13:30: Spoke3, WP3, T3.2c – Data mining, artificial intelligence, and machine learning approaches to identify subnetworks of cancer associated with early prediction, survival, metastasis or phenotypes in cancer subtypes focusing on myeloma (relatori: Simona D’Amore, Massimo Bilancia)

Abstract: Lo studio si propone di sviluppare un nuovo algoritmo prognostico/di stratificazione di progressione da MGUS a mieloma multiplo (MM) basato su Deep Diffusion Models, in grado di integrare nel modello la struttura spaziale delle componenti cellulari ed extracellulari nel microambiente del MM e di identificare i driver per la stratificazione del rischio.

Il seminario si terrà online e in lingua inglese sulla piattaforma Teams.

La partecipazione è aperta a tutti: Registrati qui.