Dennis Bontempi

Benchmarking attention-based methods for vision transformers' interpretability in retinal fundus imaging

This study benchmarks attention-based interpretability methods for Vision Transformers in retinal fundus imaging, finding gradient-weighted attention rollout provides the most faithful and biologically relevant attributions.

Sex-specific disease association and genetic architecture of retinal vascular traits

This study identifies sex-specific genetic and environmental contributions to retinal vascular traits through GWAS and interaction analyses, highlighting the need for sex-aware approaches in precision medicine.

Genetic and Etiological Insights from Automated Lumen Diameter Measurements in Carotid Ultrasounds of the UK Biobank

The paper develops a deep learning model to measure carotid lumen diameter from ultrasounds and performs GWAS, finding cLD has higher heritability and stronger associations with cardiovascular risk factors than cIMT.

Comparing tangible retinal image characteristics with deep learning features reveals their complementarity for gene association and disease prediction

CEREBRUM-7T: Fast and Fully-volumetric Brain Segmentation of 7 Tesla MR Volumes

CEREBRUM-7T is a fast, fully-volumetric brain segmentation method for 7T MRI that outperforms its training labels and is suitable for neuroimaging studies.