MUMBAI, India, June 16 -- Intellectual Property India has published a patent application (202611054529 A) filed by Kulvinder Singh; Sanjeev Dhawan; and Deepanshu, Kurukshetra, Haryana, on April 29, for 'hybrid ensemble deep learning model for accurate and robust chronic kidney disease prediction.'
Inventor(s) include Kulvinder Singh; Sanjeev Dhawan; and Deepanshu.
The application for the patent was published on June 5, under issue no. 23/2026.
According to the abstract released by the Intellectual Property India: "This current invention reveals a hybrid ensemble multi-modal deep learning system and approach to reliable and efficient prediction of Chronic Kidney Disease (CKD). The system unites disparate healthcare information as the structured clinical parameters as well as unstructured medical imaging information in a aggregated computational system. A fully connected neural network is used to handle clinical data whereas convolutional neural networks are used to handle imaging data to distinguish complementary feature representation. The extracted features are considered and their features are added together using a feature concatenation mechanism in order to create an all inclusive picture of the health status of the patient. The hybrid approach of learning which is built on stacking a set of base classifiers which may be either machine or deep learning models are used to enhance a situation with better prediction accuracy and generalization. The system also uses preprocessing methods of dealing with missing data and class imbalances, as well as optimization based on a loss reduction function. A module with features attribution and visual interpretation to improve transparency and clinical reliability will be incorporated. The suggested invention has a higher diagnostic accuracy, robustness, and scalability, and can readily be applied in the clinical decision support systems, telemedicine platforms, and intelligent healthcare application."
Disclaimer: Curated by HT Syndication.
