Understanding and predicting the progression of neurodegenerative diseases are central topics in modern neurological research. My scientific production includes a series of innovative and validated models—ranging from the pioneering DANI-Net and 4D-DANI-Net to the most recent BrLP, TADM, and their extensions—recognized and awarded by leading conferences and international journals.
Motivations & Objectives
- Simulate and predict the individual progression of neurodegeneration on longitudinal MRI data.
- Improve precision medicine by providing customizable and validated clinical tools.
- Bring generative and diffusion models from experimental research to large-scale application.
Methods
- DANI-Net/4D-DANI-Net: Adversarial networks capable of generating realistic simulations of brain pathological progression based on longitudinal MRI.
- BrLP: Spatiotemporal latent diffusion model, enhanced by prior clinical knowledge infusion, predictive consistency stabilization (LAS), and training on over 11,000 MRIs from ADNI, OASIS, AIBL.
- TADM: Diffusion-based model aimed at explicit temporal modeling of degenerative brain progression, with extensions including bidirectional temporal regularizations (TADM-3D/BiTR, arXiv 2025).
- Technical assessment on SPIE: Systematic study on comparative performance of different diffusion approaches for Alzheimer’s progression.
Results & Impact
- Significant increases in volumetric accuracy and clinical prediction compared to previous state-of-the-art.
- Awards and recognition at MICCAI, MedIA, SPIE (Best Paper nomination, Runner-up, extended publication on MedIA).
- Open-source codes and pipelines ready for reproducible testing and future applications (see repositories below).
Related Scientific Articles
- DANI-Net (MICCAI 2019):
Degenerative adversarial neuroimage nets: generating images that mimic disease progression
Authors: Daniele Ravi, Daniel C Alexander, Neil P Oxtoby, Alzheimer’s Disease Neuroimaging Initiative - 4D-DANI-Net (MedIA 2022):
Degenerative adversarial neuroimage nets for brain scan simulations: Application in ageing and dementia
Authors: Daniele Ravi, Stefano B Blumberg, Silvia Ingala, Frederik Barkhof, Daniel C Alexander, Neil P Oxtoby, Alzheimer’s Disease Neuroimaging Initiative - BrLP (MICCAI 2024 – Best Paper Nomination):
Enhancing spatiotemporal disease progression models via latent diffusion and prior knowledge
Authors: Lemuel Puglisi, Daniel C Alexander, Daniele Ravì - BrLP – Extended Version (MedIA 2025):
Brain latent progression: Individual-based spatiotemporal disease progression on 3D brain MRIs via latent diffusion
Authors: Lemuel Puglisi, Daniel C Alexander, Daniele Ravì, Alzheimer’s Disease Neuroimaging Initiative - TADM (MICCAI 2024):
Tadm: Temporally-aware diffusion model for neurodegenerative progression on brain MRI
Authors: Mattia Litrico, Francesco Guarnera, Mario Valerio Giuffrida, Daniele Ravì, Sebastiano Battiato - TADM-3D with BiTR (arXiv 2025):
Temporally-Aware Diffusion Model for Brain Progression Modelling with Bidirectional Temporal Regularisation
Authors: Mattia Litrico, Francesco Guarnera, Mario Valerio Giuffrida, Daniele Ravì, Sebastiano Battiato - Assessment on SPIE (2025):
A technical assessment of latent diffusion for Alzheimer’s disease progression
Authors: E McMaster, L Puglisi, C Gao, AR Krishnan, AM Saunders, D Ravi, T Vercauteren, DC Alexander, NP Oxtoby, Alzheimer’s Disease Neuroimaging Initiative
Code Repositories
- BrLP: LemuelPuglisi/BrLP (core:
src/brlp/) - DANI-Net & 4D-DANI-Net: daniravi/Brain-MRI-Simulator (implementations for progression, aging, and simulation)
Team & Authors
- Daniele Ravì (PI, design, supervision of DANI-Net, 4D-DANI-Net, BrLP, TADM models)
- Lemuel Puglisi (First author BrLP)
- Daniel C. Alexander, Neil P. Oxtoby, Stefano B. Blumberg, Silvia Ingala, Mattia Litrico, Francesco Guarnera, Mario Valerio Giuffrida, Sebastiano Battiato, Frederik Barkhof, E McMaster, AR Krishnan, AM Saunders, T Vercauteren, Alzheimer’s Disease Neuroimaging Initiative
Best Paper Nomination @MICCAI & MedIA.