MUMBAI, India, Aug. 24 -- Intellectual Property India has published a patent application (202621081505 A) filed by Symbiosis International Deemed University on July 01, 2026, for A Multimodal Machine Learning Framework For Prenatal Depression Detection Using Digital Phenotyping.
Inventors include Dr. Priyanka V. Deshmukh; and Dr. Aniket K. Shahade.
The application for the patent was published on August 21, 2026, under issue no. 34/2026.
Abstract: ABSTRACT A MULTIMODAL MACHINE LEARNING FRAMEWORK FOR PRENATAL DEPRESSION DETECTION USING DIGITAL PHENOTYPING The present invention discloses a multimodal framework (100) for prenatal depression detection integrating smartphone sensors (112), voice analysis (114), and wearable physiological data (116) with a pregnancy-specific augmentation module (120) adapting physiological (122), behavioral (124), and vocal (126) parameters for gestational changes. A multimodal feature engineering module (130) extracts trimester-adjusted biomarkers. A hybrid fusion module (140) combines early (142), late (144), and cross-modal attention (146) strategies. A trimester adaptation mechanism (152) adjusts classification thresholds. The framework achieves 84.3 percent accuracy with 0.89 AUC, improving detection by 14.7 percent over non-adapted models. [
Disclaimer: Curated by HT Syndication.
