Researchers have developed DeMMO, a novel interpretable framework designed for longitudinal, multi-disease, and multi-outcome learning using digital mobility outcomes (DMOs) from wearable sensors. This framework addresses limitations in existing temporal multi-task models by enabling joint modeling of multiple prediction outcomes across different diseases, even when patient cohorts do not overlap. DeMMO achieves superior prediction performance on the Mobilise-D dataset, outperforming nine established baselines, and includes a mechanism for automatic cross-disease and cross-outcome relation learning. AI
IMPACT This framework could improve disease progression monitoring and clinical validation through advanced analysis of wearable sensor data.
RANK_REASON The cluster contains a research paper detailing a new machine learning framework. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
