Pupillary curve morphology and diagnosis management
Inventors
McGrath, Lynn B. • Maxin, Anthony • Alfieri, Robin
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-12758157-B2
Publication Date
2026-10-13
Expiration Date
Abstract
Techniques for calculating or managing pupillary curve data identified based on data associated with computing devices are discussed herein. For example, a machine learning (ML) model can be utilized to analyze sensor data collected by a camera associated with a computing device. The ML model can perform a comparison between a pupillary (e.g., a pupil response curve and previous pupillary curves (e.g., previous pupil response curves) utilizing classification information associated with the previous pupil response curves. The comparison can be utilized to identify a physiological condition associated with the pupil response curve. Information identifying the physiological condition can be presented by a display or transmitted to the computing device.
Core Innovation
The disclosed method and system capture sensor data associated with pupillary activity and represent changes in pupil characteristics over time as pupillary activity curves. They generate classifications associated with physiological conditions and baselines, train a machine learning model using library data, and analyze a current curve generated from a simultaneous binocular scan. The result is output as a physiological condition identifier.
The disclosed analysis processes whole pupillary activity curves to identify patterns and characteristics without computing intermediate pupillary metrics prior to classification. The supplied partial content also describes camera and video data, light or other stimuli, and predictive measurements. It states that the systems reduce compute, network, and memory use compared with conventional pupillometry systems. The background section is not included in the supplied content, so the problem described there is not explicitly stated.
Claims Coverage
The independent claims are claims 1, 8, and 16, which recite three inventive features across a method, a processor-based system, and computer-readable media storing executable instructions.
Tagged curve library and binocular comparison method
Claim 1 monitors pupils, captures sensor data, and collects pupillary activity curves in database files. It generates classification models from whole curves and tagged library data, generates tagged baselines representing known physiological curves without at least one physiological condition, and trains a machine learning model. The method then compares a current curve representing both pupils during a simultaneous binocular scan with the library curves and outputs a physiological condition identifier.
Processor-based system for analyzing binocular pupillary activity
Claim 8 covers processors and computer-readable media storing instructions that capture sensor data, collect datasets containing pupillary activity curves, generate classifications and baselines, and analyze a current curve representing both pupils during a simultaneous binocular scan. A machine learning model trained based on the baselines and classifications analyzes the current curve, and the system outputs a physiological condition identifier.
Computer-readable media for curve-based condition identification
Claim 16 covers non-transitory computer-readable media storing instructions that cause a processor to capture sensor data, collect datasets containing pupillary activity curves, and generate a current curve from simultaneous binocular data. One or more data algorithm models analyze the current curve based on the pupillary activity curves, and the result is output as a physiological condition identifier.
The independent claims cover collecting and representing pupillary activity as curves, analyzing current binocular pupil responses using stored curves and, where specified, classifications and baselines, and outputting a physiological condition identifier. They claim these functions as a method, a processor-based system, and processor-executable instructions.
Stated Advantages
The disclosure states that the systems reduce compute, network, and memory use compared with conventional pupillometry systems.
Documented Applications
The claims describe analyzing pupillary activity to identify physiological conditions. Listed examples include injury, intoxication, deception, fatigue, dementia, disease, heart condition, and diabetes.
A dependent claim describes outputting a predictive performance measurement indicating the probability of future coma or vegetative-state likelihood for an individual.
Interested in licensing this patent?