MUMBAI, India, June 26 -- Intellectual Property India has published a patent application (202641072734 A) filed by Jagdish Devarajan on June 11, 2026, for A Computer-Implemented System And Method For Dynamic Minimal Clinical Signal Selection And Time-Granular Patient Trajectory Validation.
Inventor includes Jagdish Devarajan.
The application for the patent was published on June 19, 2026, under issue no. 25/2026.
Abstract: A computer-implemented clinical information processing system determines, for each milestone of a guideline-derived expected treatment trajectory, a minimal clinically sufficient signal set and uses that set as a gate on computation. While the collected minimal signal set confirms the patient is on the expected milestone, the system suppresses execution of artificial-intelligence models on the patient record and records a confirmation event. Only when the minimal signal set deviates from an expected threshold, or an uncertainty measure is exceeded, does the system select the least-costly sufficient inference mode from a tiered set of deterministic rules, small local, condition-specific and large models. A resource-footprint ledger quantifies the resulting reduction in processor, memory, storage, network and model- execution use relative to processing a larger portion of the record. In embodiments the system withholds subsequent milestones, branches unrelated signals, tracks expected treatment effects as sub-streams, and compares the trajectory against a once-generated natural-history model.
Disclaimer: Curated by HT Syndication.
