MUMBAI, India, Sept. 14 -- Intellectual Property India has published a patent application (202641103236 A) filed by Dr. Murugananth Gopal Raj on August 27, 2026, for A Hybrid Quantum-Classical Convolutional Neural Network System For Mri-Based Brain Tumor Image Processing.
Inventors include Dr. Reji R; and Mrs. S. Varaunadevi.
The application for the patent was published on September 11, 2026, under issue no. 37/2026.
Abstract: A hybrid quantum-classical convolutional neural network system and method are disclosed for automated brain tumor detection from MRI data. An MRI input module supplies preprocessed image data to a ResNet152-based classical feature extraction module, generating hierarchical feature representations that are projected into a four- dimensional vector for quantum encoding. A four-qubit 10 parameterized quantum circuit applies RX-RZ-RX rotation gates, followed by controlled-RX and controlled- Z entanglement operations arranged in configurable topologies, with data re-uploading and quantum pooling enhancing feature expressiveness. Measurement outputs drive a classification module trained jointly via Adam and SPSA optimizers. A noise-aware simulation module evaluates circuit resilience under simulated quantum noise before deployment on physical quantum processors (IBM Q, 15 lonQ), applying error-mitigation techniques for reliable execution. A Grad-CAM explainability module and DICOM/PACS-compatible interface enable radiologist-interpretable, clinically integrated outputs. The invention improves diagnostic accuracy, generalization, and computational efficiency over conventional classical-only approaches while remaining compatible with near-term quantum hardware.
Disclaimer: Curated by HT Syndication.
