Medical AI | Wenrui Fan

Foundation-model-boosted multimodal learning for fMRI-based neuropathic pain drug response prediction

Mon, 01 Sep 2025 00:00:00 +0000

We study how foundation models can transfer knowledge from larger, pain-agnostic datasets into small, pain-specific neuroimaging cohorts for drug-response prediction.

Foundation models for data-scarce neuropathic pain research

Mon, 03 Mar 2025 12:00:00 +0000

Neuropathic pain is difficult to treat, and data scarcity makes it challenging to train high-capacity models for drug-response prediction. Our work studies how foundation models can transfer knowledge from larger, pain-agnostic datasets into small, pain-specific neuroimaging cohorts.

The resulting framework combines multimodal information within the target cohort with representations learned from external fMRI data. The paper is available on .

Multimodal variational autoencoder for low-cost cardiac hemodynamics instability detection

Sun, 06 Oct 2024 00:00:00 +0000

CardioVAE released

Wed, 20 Mar 2024 12:00:00 +0000

, our multimodal approach to low-cost cardiac hemodynamics assessment, is available on arXiv.

MeDSLIP released

Fri, 15 Mar 2024 12:00:00 +0000

We released , a dual-stream vision-language pre-training framework for fine-grained medical image-text alignment.

MeDSLIP: medical dual-stream language-image pre-training with pathology-anatomy semantic alignment

Fri, 15 Mar 2024 00:00:00 +0000

Foundation Models for Neuropathic Pain

Mon, 01 Jan 0001 00:00:00 +0000

Neuropathic pain is difficult to treat, while therapeutic response varies substantially across patients. This project combines foundation models and multimodal representation learning to make better use of small neuroimaging cohorts and predict response to treatment.

Multimodal AI for Parkinson's Disease

Mon, 01 Jan 0001 00:00:00 +0000

This interdisciplinary project develops multimodal AI methods to improve understanding, diagnosis, and prognosis of Parkinson’s disease. It brings together the Sheffield Institute for Translational Neuroscience and the Faculty of Engineering at the University of Sheffield.