MeditronFO (Fully Open) is the first fully open medical specialist LLM, and outperforms Medgemma on open-ended clinical evaluations.

MoBayes constructs an explicit clinical knowledge base, then iteratively gathers evidence through LLM-parsed patient dialogue and updates beliefs with Bayesian inference — outperforming much larger standalone LLM doctors at a fraction of the cost. The LLM is confined to parsing and verbalization, while a deterministic Bayesian module handles posterior tracking, question selection, and calibrated abstention.

MultiModN explores modular neural architectures for multimodal, multi-task learning, with a focus on interpretable fusion across heterogeneous data sources.
The project introduces a flexible approach to combining modalities sequentially, supporting robust prediction even when some information is missing not at random.

DeepBreath studies how deep learning can detect pediatric respiratory pathology from lung auscultation audio, aiming to make respiratory assessment more objective and scalable. The project highlights interpretable audio-based signatures of disease, with potential relevance for standardized evaluation in remote and resource-limited care settings.

Asthmoscope investigates whether deep learning on digital lung auscultation can support pediatric asthma homecare monitoring, using spirometry-labelled recordings as an objective clinical reference.
The project explores both the promise and current limitations of audio-based asthma assessment, highlighting challenges such as early exacerbation detection and variability across digital stethoscopes.