MUMBAI, India, Aug. 12 -- Intellectual Property India has published a patent application (202621075885 A) filed by Symbiosis International Deemed University on June 18, 2026, for System And Method For Predicting Autoimmune Diseases By Machine Learning Analysis Of Sputum Biomarkers.
Inventor includes Dr. Gaurav Vishnu Londhe.
The application for the patent was published on August 07, 2026, under issue no. 32/2026.
Abstract: ABSTRACT SYSTEM AND METHOD FOR PREDICTING AUTOIMMUNE DISEASES BY MACHINE LEARNING ANALYSIS OF SPUTUM BIOMARKERS The present invention discloses a non-invasive system (100) and method for predicting autoimmune diseases from sputum. A sputum collection and processing module (110) prepares a sample from which a multimodal data acquisition and feature extraction module (120) derives cellular, molecular, genomic, proteomic, metabolomic, and microbiome features. A data preprocessing module (130) and a feature engineering module (140) clean, harmonize, and reduce the features into a fused representation. A machine learning and deep learning module (150) applies classical and deep learning models, including autoencoders and temporal networks, to capture non-linear immune-dysregulation patterns. A prediction and decision module (160) outputs disease classification, subtype identification, severity stratification, treatment response prediction, and flare-versus-remission detection. An explainability and visualization module (170) attributes each prediction to contributing biomarkers, and a clinical integration module (180) presents results through an AI-assisted dashboard, while federated training preserves patient privacy, enabling early, accurate, and clinically interpretable diagnosis. [
Disclaimer: Curated by HT Syndication.
