Non-invasive non-contact system and method for measuring diabetes mellitus condition using thermal imaging
Inventors
Shivpure, Sameer Raghuram • THIRUVENGADAM, Jayanthi • Choda, Anuhya • CHODA, Gayathri
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-12023130-B2
Publication Date
2024-07-02
Expiration Date
Abstract
System and method for measuring diabetes mellitus condition of a subject is disclosed. The disclosed system and method includes thermal sensors for capturing thermal images and/or videos of a body part; and a processing engine to detect a predefined region of the body part in each frame of the captured images and/or videos. The processing engine segments one or more portions from the detected predefined region in each frame of the captured images and/or videos to identify a region of interest comprising major arteries in the segmented portions. Based on the ROI, the engine extracts pixel values, representing biosignals, from each frame of the captured images and/or videos so as to determine one or more parameters associated with the hemodynamic factors and a rate of atherosclerosis of the subject. Further, a risk score for the diabetes mellitus condition based on the determined parameters using computational models is measured.
Core Innovation
The invention relates to a non-contact, non-invasive thermal imaging system for measuring diabetes mellitus condition of a subject. A set of thermal sensors captures thermal images and videos of at least one body part, and a processing engine receives data packets associated with the captured frames. The processing engine detects a predefined region in each frame, segments one or more portions from the detected predefined region, and identifies a region of interest comprising one or more arteries in the segmented portions.
From the identified region of interest in each frame, the processing engine extracts one or more pixel values representing a set of biosignals. The processing engine determines one or more parameters associated with hemodynamic factors and a rate of atherosclerosis based on the extracted pixel values. Based on the determined one or more parameters, the processing engine measures a risk score for the diabetes mellitus condition.
The system measures the risk score using computational models and generates the computational models by calculating signal parameters for one or more signals associated with the one or more arteries in both a healthy subject and a diabetes subject. The processing engine identifies a set of parameters that correlate to complications associated with the diabetes condition using principal component analysis on the calculated signal parameters, and determines patterns and differences among parameters between diabetes and healthy subjects using statistical methods and visualization. The processing engine trains the computational models using machine learning units comprising clustering models, logistic regression, random forest, or neural network models on the parameters.
Claims Coverage
The document includes two independent claims, each covering a thermal-imaging pipeline that derives a diabetes mellitus risk score from artery-related hemodynamic and atherosclerosis parameters using computational models trained with machine learning; the second claim recites the corresponding method steps.
Thermal imaging system for diabetes risk scoring from arterial biosignals
A system for measuring diabetes mellitus condition comprising a set of thermal sensors capturing thermal images and videos of a subject, and a processing engine that detects a predefined region, segments portions, identifies a region of interest comprising one or more arteries, extracts pixel values representing a set of biosignals, determines parameters associated with hemodynamic factors and a rate of atherosclerosis, and measures a risk score using computational models trained via signal-parameter calculation for healthy and diabetes subjects, principal component analysis, statistical methods and visualization, and machine learning units comprising clustering models, logistic regression, random forest, or neural network models.
Method for measuring diabetes risk score using thermal imaging and computational model training
A method comprising capturing thermal images and videos of at least one body part, receiving data packets, detecting a predefined region, segmenting portions, identifying a region of interest comprising one or more arteries, extracting pixel values representing a set of biosignals, determining parameters associated with hemodynamic factors and a rate of atherosclerosis, and measuring a risk score using computational models, wherein generating the computational models includes calculating signal parameters for one or more signals associated with the one or more arteries in both a healthy subject and a diabetes subject, identifying parameters correlated to diabetes complications using principal component analysis, determining patterns and differences using statistical methods and visualization, and training the computational models using machine learning units comprising clustering models, logistic regression, random forest, or neural network models on the parameters.
Across the independent claims, inventive coverage centers on thermal sensor capture, predefined-region detection, segmentation, artery-based region-of-interest extraction of pixel values as biosignals, determination of hemodynamic and atherosclerosis-rate parameters, and computation of a diabetes mellitus risk score using computational models generated with PCA, statistical methods and visualization, and machine-learning training on artery-associated signal parameters from healthy and diabetes subjects.
Stated Advantages
Early/provisional diagnosis of diabetes mellitus condition.
Monitoring progression and treatment efficacy.
Use of vascular/hemodynamic biomarkers.
Avoidance of harmful radiation due to non-contact, non-invasive thermal imaging.
Documented Applications
Measuring diabetes mellitus condition and generating a diabetes risk score from non-contact, non-invasive thermal imaging.
Monitoring progression and treatment efficacy of diabetes mellitus condition using derived parameters associated with hemodynamic factors and a rate of atherosclerosis.
Assessing diabetes complications associated with diabetes condition using parameters correlated to complications identified via principal component analysis and modeling.
Interested in licensing this patent?