10:20-13:00

(10 presentations, 15 min each — 2 h 30 min + 10 min overhead)

ORAL SESSION I — Explainable & Trustworthy AI 

Measuring Explainable AI: A Quantitative Analysis of XAI Methods for Melanoma Classification — Nath, John, Iacca

Developing a Semantic and Explainable Artificial Intelligence System to Improve Clinical Prioritization Within Homogeneous Waiting Groups (HWG) in the Italian Healthcare System — Tombolini, Ponte, Mariotti, Dragoni

Efficient Ensemble Approaches for Explainable Artificial Intelligence (XAI) in Image Classification — Shafi, Iacca

XLUS-ViT: An Explainable Vision Transformer Approach for COVID-19 Lung Ultrasound Classification — Khalid, Iacca

Ethics-by-Design Framework to Support Trustworthy AI in Predicting Prediabetes: The PRAESIIDIUM Case Study — Masetti, Operto, Paglialonga, Gianfreda, Veruggio, PRAESIIDIUM Consortium

SESSION CHAIRS: TOMASSINI, ROSANI

ORAL SESSION II — Neuroimaging & Disease Modeling

Exploiting the Rashomon Set: Toward Stable and Interpretable MRI Biomarkers for Alzheimer’s Disease — Borsani, Rosani, Abdi, Di Fatta

Low-Field MRI-Based Deep Segmentation of Perivascular Spaces in Parkinson’s Disease — Tomassini, Girardi, Di Giacopo, Quattrocchi, Giorgini

Benchmarking Radiomics, Deep Learning, and Foundation Models for Glioblastoma Survival Prediction — Marasi, Doniselli, Pascuzzo, Aquino, Redaelli, De Momi, Marzullo

Small-scale Disease Progression Model of Glioblastoma via Physics Informed Neural Network — Vinci

Adapting to the Inevitable: Lightweight Volume-Wise Fine-Tuning for Out-of-Distribution Brain Tumor Segmentation — Chaudhry, Tomassini, Giorgini

* Prognostic Modeling in Glioblastoma Using WSI Foundation Models and Transcriptomic Profiles — Rignanese, Sabbatini, Barone, Noei, Pozzi, Manganaro, Ragni, Bovo, Novello, Chierici, Falco, Jurman

SESSION CHAIRS: TOMASSINI, RAIMI

Presentation marked with * will be delivered remotely at the end of the last oral session of the afternoon (approximately 17:00 onwards)

15:00-17:40 

(10 presentations, 15 min each — 2 h 30 min + 10 min overhead)

ORAL SESSION III — Data Integration & Clinical Decision Support

A Topological Data Analysis–Informed Machine Learning Framework for Predicting Recovery After Acute Ischemic Stroke Thrombectomy — Rucco, Viticchi, Falsetti, Silvestrini, Polonara

AI-Based Personalization of Diabetic Retinopathy Screening Intervals Through Multimodal Integration — Bresolin, Cocu, Malfatti, Ragni, Bovo, Cagol, Inchiostro, Romanelli, Moroni, Jurman

AI-Based Diagnosis of Colonic Polypoid Lesions From Endoscopic Biopsies — Pozzi, Noei, Zordan, Pegoraro, Bovo, Ragni, Novello, Barbareschi, Jurman

Is Self-Pretraining Really Useful for Medical Time Series? A Transformer-Based Study — Coser, Orvieto, Soda, Zollo

Exploring Acoustic Signatures of Vocal Pathologies: Foundation-Model Speech Representations for Malignant Voice Disorder Detection — Salvaterra, Zen, Oss Emer, Barone, Grosso, Demattè, Perotti, Matassoni, Brutti, Tessadori, Novello, Ioppi, Piccin, Jurman

SESSION CHAIRS: GIRARDI, MARZULLO

ORAL SESSION IV — Robotics & Agentic AI 

An Anatomy-Aware Shared Control Approach for Assisted Teleoperation of Lung Ultrasound Examinations — Nardi, Fontanelli, Saveriano, Palopoli, Lamon

Autonomous Robotic Palpation and Abnormality Detection Through Ergodic Exploration — Beber, Lamon, Saveriano, Fontanelli, Palopoli

An Agentic System for Data Harmonization and Augmentation in Clinical Trials — Marzullo, De Momi

* Automating Surgical Debriefing Through Agentic AI — Fumi, Bombieri, Allievi, Bonvini, Chaspari, Zenati, Giorgini

SESSION CHAIRS: TOMASSINI, LAMON

Presentation marked with * will be delivered remotely (approximately 17:00 onwards)