MUMBAI, India, Oct. 12 -- Intellectual Property India has published a patent application (202641118998 A) filed by Vallurupalli Nageswara Rao Vignana Jyothi Institute Of Engineering And Technology on October 05, 2026, for A Quantum-Enhanced Generative Model System And Method For Brain Tumor Analysis Using Amplitude Estimation.
Inventor includes Dr. G. Madhu.
The application for the patent was published on October 09, 2026, under issue no. 41/2026.
Abstract: ABSTRACT [0018] A hybrid quantum-classical system is disclosed for generating multiple plausible segmentations of brain tumors from multi-modal MRI volumes and for efficiently estimating the probabilities of clinically relevant tumor properties together with associated uncertainty levels. The system employs a Conditional Variational Autoencoder trained on stacked multi-modal MRI inputs to learn a structured latent distribution over tumor shapes. Sampling from this latent space produces an ensemble of segmentation hypotheses, referred to as dreams, that capture anatomical variability and boundary ambiguity. These dreams are encoded into a quantum superposition state. A quantum oracle marks those dreams that satisfy a chosen clinical criterion, such as multifocality. Quantum Amplitude Estimation then recovers the probability of the marked property with quadratic improvement in query complexity relative to classical sampling. The resulting probability, together with derived statistics such as mean tumor volume and inter-dream variability, is presented in a clinical dashboard. The approach yields both pixel-level segmentation masks and high-level probabilistic clinical metrics while substantially reducing the number of model evaluations required for reliable uncertainty quantification. The system is trained and validated on the BraTS multi-modal MRI dataset and is readily extensible to other volumetric medical imaging modalities.
Disclaimer: Curated by HT Syndication.
