MUMBAI, India, April 17 -- Intellectual Property India has published a patent application (202641018835 A) filed by Nandha Engineering College, Erode, Tamil Nadu, on Feb. 19, for 'non-invasive anemia detection using convnext-based deep learning with explainable visual analysis.'
Inventor(s) include R Somambika; M Shrivarshini; A Sowmya; and B Deepa.
The application for the patent was published on April 17, under issue no. 16/2026.
According to the abstract released by the Intellectual Property India: "The present invention relates to a non-invasi, accurate, and explainable system for anemia detection using deep learning based medical image analysis. Anemia is a widespread health condition traditionally diagnosed through invasive blood tests, which arc time-consuming, costly, and often inaccessible in resource-limited settings. Visual examination of the conjunctival region of the eye provides a clinically relevant indicator of anemia; however, manual assessment is subjective and prone to variability. Existing automated approaches either rely on handcrafted features or lack interpretability, limiting their clinical reliability. The proposed invention addresses these limitations by introducing a ConvNeXt based deep convolutional neural network for automated analysis of conjunctival images. The system utilizes contrast enhancement techniques to amplify subtle chromatic and texture variations associated with hemoglobin deficiency, enabling robust feature extraction from high-resolution ocular images. The ConvNeXt architecture combines the efficiency of convolutional networks with modern design principks, providing high classification accuracy while remaining computationally efficient. To ensure transparency and clinical trust, the invention incorporates explainable artificial intelligence (XAI) techniques, including Grad-CAM and Integrated Gradients, to visualize regions of the conjunctiva that. contribute most to the model's decision. These explanations allow healthcare professionals to validate predictions and better understand the underlying visual cues used by the system. The proposed framework is lightweight, scalable, and suitable for real-time deployment on standard computing devices. By enabling accurate, interpretable, and non-invasive anemia screening, the invention provides a practical and cost-effective solution for early diagnosis and large-scale population screening, thereby improving accessibility and reliability of anemia detection in clinical and community healthcare settings."
Disclaimer: Curated by HT Syndication.
