MUMBAI, India, July 24 -- Intellectual Property India has published a patent application (202641085604 A) filed by Vallurupalli Nageswara Rao Vignana Jyothi Institute Of Engineering And Technology on July 13, 2026, for Ai-Based Brain Tumour Detection And Classification Framework Using 3d-Unet And Transformer Networks.
Inventors include Dr. Kriti Ohri; and Dr. R. Vijaya Saraswati.
The application for the patent was published on July 17, 2026, under issue no. 29/2026.
Abstract: Early detection and diagnosis of brain tumours are critical for improving medical treatment outcomes and increasing a patient’s chances of survival. Brain tumours, if left undetected or misdiagnosed, can lead to severe neurological impairments or life-threatening conditions. Magnetic Resonance Imaging (MRI) is a widely used imaging technique that provides high-resolution scans of brain tissues, allowing for accurate identification of abnormalities. However, manually analysing MRI images to detect tumours is a complex, time-intensive task that heavily depends on the expertise and experience of medical professionals. This manual approach is prone to human error and variability in diagnosis, which can lead to delays in treatment. To overcome these challenges, the development of an automated framework for brain tumour segmentation and classification is essential. By integrating artificial intelligence (AI) techniques, particularly deep learning models, this framework can significantly improve accuracy, consistency, and efficiency in tumour detection.A robust framework for brain tumour detection should leverage advanced deep learning techniques to enhance both segmentation and classification accuracy. Specifically, the proposed approach should incorporate a 3d-UNet model, which is effective for volumetric segmentation of MRI images. This model enables precise identification and localisation of tumour regions by capturing spatial information across multiple layers of the scan. Additionally, transformer-based models, known for their ability to analyse complex patterns and contextual relationships, should be employed for tumour classification. By utilising transformers, the framework can distinguish between different tumour types with improved accuracy, ensuring a more reliable diagnosis. The combination of these techniques will create a highly efficient system capable of handling large datasets while maintaining precision in tumour identification.To ensure the effectiveness of the proposed framework, rigorous validation and performance evaluation must be conducted. This process should involve the implementation of key assessment metrics such as loss functions and precision-recall diagrams to measure the accuracy and reliability of the model. Furthermore, the developed system should be compared against existing models to assess its advantages in terms of efficiency, scalability, and diagnostic accuracy. The goal is to provide a more reliable and automated diagnostic tool that reduces the dependency on manual analysis while supporting medical professionals in making informed treatment decisions. By integrating AI-driven techniques, the framework has the potential to revolutionise brain tumour detection, leading to faster diagnosis and improved patient outcomes.
Disclaimer: Curated by HT Syndication.
