MUMBAI, India, Sept. 22 -- Intellectual Property India has published a patent application (202641110084 A) filed by Vellore Institute Of Technology on September 14, 2026, for Method For Predicting Cancer Risk From Dietary Habits And Gut Microbiome Using A Feedforward Neural Network.
Inventors include Sudha M; and Harsh Kumar.
The application for the patent was published on September 18, 2026, under issue no. 38/2026.
Abstract: ABSTRACT METHOD FOR PREDICTING CANCER RISK FROM DIETARY HABITS AND GUT MICROBIOME USING A FEEDFORWARD NEURAL NETWORK A computer-implemented method of predicting cancer risk for a subject obtains gut-microbiome-derived data (102), computes at least one microbial diversity metric (104), and supplies the microbial diversity metric (104) to a trained feedforward neural network (200). The method assembles a multimodal feature vector (110) jointly containing dietary intake quantities (106), the microbial diversity metric (104), a short-chain fatty acid concentration (108), an inflammatory biomarker level (112), and a demographic attribute (114). The multimodal feature vector (110) is scaled to a common numerical range by a fitted scaling transformation (120) to produce a scaled feature vector (122). The scaled feature vector (122) is applied to the feedforward neural network (200) trained on class-weighted samples to output a binary risk classification (304) indicating whether the subject is at high risk or low risk for cancer. (Fig. 1)
Disclaimer: Curated by HT Syndication.
