Predicting intraoperative cardiac arrest using Temporal Attention Networks on VitalDB clinical time-series data
Fed-CA: complementarity-aware federated aggregation for deep time-series imputation (SAITS) under heterogeneous MNAR, validated on VitalDB clinical data.
HRV/ML pipeline predicting intraoperative arrhythmia onset from the public VitalDB Arrhythmia Database annotation stream
Intra-operative hypotension prediction
AnestheSIM is a clinical-data-integrated patient simulator for anesthesia modeling control. It reproduces realistic patient physiology from high-resolution clinical data, enabling the development and benchmarking of automated drug-delivery and depth-of-anesthesia control algorithms.
Multi-event intraoperative early-warning on VitalDB — live demo + honest benchmark. Research prototype.
Multi-module clinical ML framework for real-time ICU mortality prediction, multimorbidity risk scoring, and waveform event detection with automated MLOps drift monitoring and ensemble fusion.
ReliabiliReliability-aware self-correcting digital twin for clinical vital sign validation and health risk assessment using GRU-based deep learning.
Two-cohort benchmark of zero-shot time-series foundation models vs. task-trained models for intraoperative hypotension prediction (MAP forecasting on VitalDB & MOVER).
To associate your repository with the vitaldb topic, visit your repo's landing page and select "manage topics."