MUMBAI, India, Sept. 22 -- Intellectual Property India has published a patent application (202641109851 A) filed by Vallurupalli Nageswara Rao Vignana Jyothi Institute Of Engineering And Technology on September 13, 2026, for A Hybrid Deep Learning System For Automated Brain Tumor Detection In Fetal And Infant Medical Images.
Inventor includes Mr. Girish Kulkarni.
The application for the patent was published on September 18, 2026, under issue no. 38/2026.
Abstract: ABSTRACT [0016] An artificial-intelligence system for early detection of brain tumors in infants is disclosed that fuses real-time ultrasound with high-resolution MRI data and processes the combined information through convolutional and recurrent neural networks. Preprocessing steps including noise reduction, intensity normalization, skull stripping, binarization, and median filtering prepare the images. Discrete wavelet transform extracts features that are reduced by principal component analysis and classified by a kernel convolutional neural network, achieving approximately 97 percent accuracy in distinguishing tumor types. Automated segmentation delineates tumor boundaries to support treatment planning. Data augmentation and transfer learning compensate for the scarcity of annotated pediatric datasets. The resulting platform reduces operator dependency, accelerates diagnosis, and improves consistency in resource-limited clinical settings.
Disclaimer: Curated by HT Syndication.
