MUMBAI, India, Feb. 13 -- Intellectual Property India has published a patent application (202641009092 A) filed by Vellore Institute Of Technology, Chennai, Tamil Nadu, on Jan. 29, for 'system and method to evaluate and classify medical images.'
Inventor(s) include K. Lekshmi; Farhaan Areeb; and Divyansh Vashist.
The application for the patent was published on Feb. 13, under issue no. 07/2026.
According to the abstract released by the Intellectual Property India: "The present disclosure discloses a system (102) and a method (200) for evaluating and classifying a medical image. The system (100) includes a user interface (102) that receives computed tomography (CT) images from sources. A processor (104) connected to the user interface (102) to receive medical images, including real CT image and synthetic CT image. The processor (104) extracts feature representations from real CT images and synthetic CT images using a Siamese neural network model (104-1). Attended feature representations are generated by applying cross-attention blocks between the extracted feature representations. An adaptive similarity representation is computed using similarity operations. The adaptive similarity representation is processed using multi-layer perceptron mechanism (104-2) and sigmoid activation function (104-3) to generate a similarity score. Based on similarity score, processor (104) determines whether the synthetic CT image corresponds to memorized image or non-memorized image with respect to generative artificial intelligence model (104-4)."
Disclaimer: Curated by HT Syndication.
