MUMBAI, India, Aug. 17 -- Intellectual Property India has published a patent application (202631059604 A) filed by Sandip Dutta; Sagnik Paul; Srijan Das; Anirban Dey; Deepam Gangopadhyay; Harinandan Tunga; and Samarjit Kar on May 11, 2026, for Sasthoseva: Non-Invasive Cholesterol Estimation System Using Optical Biosensing With Multi-Modal Ai.
Inventors include Sandip Dutta; Sagnik Paul; Srijan Das; Anirban Dey; Deepam Gangopadhyay; Harinandan Tunga; and Samarjit Kar.
The application for the patent was published on August 14, 2026, under issue no. 33/2026.
Abstract: ABSTRACT: This patent application SasthoSeva introduces a multimodal non-invasive optical technology for estimating cholesterol risk without requiring blood samples. The device integrates Near-Infrared Diffuse Reflectance Spectroscopy (NIR DRS) using lipid-selective wavelengths, Photoplethysmography (PPG) for measuring the arterial stiffness along with the waveform features, Skin Autofluorescence (SAF) for measuring the Advanced Glycation End-products (AGEs), and Bioimpedance sensor for hydration and signal normalization. These technologies collectively form a 22-dimensional feature vector used for risk estimation through machine learning algorithm, XGBoost (optionally supplemented by 1D-CNN for waveform features), producing total cholesterol range, risk level (Normal, Borderline, or High), and a 90% confidence interval using SHAP-based feature importance. Longitudinal trends are also analyzed to identify progressive risk increases. Customized for the Indian population, considering skin color and demographic variations, the technology provides quick ( 60 sec.), inexpensive, and painless estimation of dyslipidaemia risk as well as continuous monitoring complementing the lab-based lipid testing.
Disclaimer: Curated by HT Syndication.
