MUMBAI, India, Aug. 31 -- Intellectual Property India has published a patent application (202641101723 A) filed by J. J College Of Engineering And Technology on August 23, 2026, for Artificial Intelligence-Based Multi-Modal Early Sepsis Prediction And Clinical Decision Support System Using Real-Time Physiological Data Fusion.
Inventors include Suganya. P; Yobu D; Mrs M Saranya; Sathiya Priya T; Mrs A Apoorvavalli; Ms Rajeshwari S; Abinaya. N; and Mrs Mahalakshmi P.
The application for the patent was published on August 28, 2026, under issue no. 35/2026.
Abstract: The present invention discloses an Artificial Intelligence-based multi-modal early sepsis prediction and clinical decision support system configured to continuously analyze heterogeneous real-time physiological and clinical information for early identification of patients at elevated risk of sepsis. The system comprises physiological sensors, clinical data integration, signal quality assessment, preprocessing, temporal synchronization, patient-specific baseline modeling, multi-modal feature extraction, an artificial intelligence prediction engine, a multi-modal fusion engine, confidence estimation, explainable artificial intelligence, and a clinical decision- support and alert module. The physiological data may include heart rate, respiratory rate, blood pressure, oxygen saturation, body temperature, ECG and other physiological parameters, while clinical information may include laboratory measurements, electronic health records, demographic information, medications and clinical observations. The system synchronizes asynchronous data streams, detects signal-quality issues, extracts temporal and patient-specific deviation features, and generates modality-specific representations. An AI-based fusion mechanism combines the representations using data-quality and relevance-dependent weighting to produce a continuously updated sepsis-risk prediction. A confidence module estimates prediction uncertainty, while an explainability module identifies major physiological and clinical factors contributing to the prediction. A clinical decision-support module generates prioritized alerts and provides risk level, confidence, temporal trends and contributing factors to authorized healthcare professionals. The system is configured to complement clinical assessment and established sepsis management protocols and does not autonomously diagnose or treat the patient. The invention facilitates continuous, personalized, multimodal and interpretable early warning for supporting timely clinical evaluation of potential sepsis.
Disclaimer: Curated by HT Syndication.
