MUMBAI, India, Sept. 28 -- Intellectual Property India has published a patent application (202641112720 A) filed by Dr. Suresh H; Bimba Prasad; Dr. M Shilpa; Dr. Nanda Ashwin; Nikitha V P; Dr. Rohini A; and Dr. Chandrakala B M on September 20, 2026, for System And Method For Real-Time Predictive Risk Stratification In Patient Monitoring Using Neural Networks.
Inventors include Dr. Suresh H; Bimba Prasad; Dr. M Shilpa; Dr. Nanda Ashwin; Nikitha V P; Dr. Rohini A; and Dr. Chandrakala B M.
The application for the patent was published on September 25, 2026, under issue no. 39/2026.
Abstract: A system and computer-implemented method are disclosed for real-time predictive risk stratification in patient monitoring. Time-stamped physiological data streams and optional clinical context are quality-assessed, normalized and temporally aligned to form successive patient-data windows. A patient-specific baseline module generates deviation information and selectively adapts the baseline under controlled stability conditions. A neural risk inference engine processes the temporal representation to generate predictive risk values for one or more future prediction horizons. An uncertainty module determines confidence using signal quality, missingness, temporal staleness, model disagreement or out-of-distribution information. A risk stratification module jointly evaluates risk magnitude, trend, confidence and persistence and applies hysteresis to control escalation and de-escalation. When escalation criteria are satisfied, an alert containing the risk state, prediction horizon, confidence and contributing factors is generated and recorded with model and audit information. The architecture thereby provides patient-specific, uncertainty-aware and persistence-controlled predictive monitoring. Accompanied Drawing [FIGS. 1-2].
Disclaimer: Curated by HT Syndication.
