MUMBAI, India, Sept. 28 -- Intellectual Property India has published a patent application (202611046296 A) filed by Dr Br Ambedkar National Institute Of Technology on April 10, 2026, for A Lightweight Method For Non-Invasive Anemia Classification Using Smartphones.
Inventors include Dr. Avani Vyas; Dr. Kusum Bharti; Dr. Nisha Chaurasia; and Dr. Ranjeet Kumar Rout.
The application for the patent was published on September 25, 2026, under issue no. 39/2026.
Abstract: The present invention (100) relates to the field of medical data processing and embedded deep learning systems. It addresses the limitations of conventional invasive techniques by providing a smartphone-based, lightweight computational framework for analyzing conjunctival eye images. The invention comprises preprocessing of images in CIELAB color space (101), feature extraction using MobileNetV1 with depthwise separable convolutions (102), and integration of Squeeze-and-Excitation (SE) blocks (103) for channel recalibration. A dual-branch architecture fuses visual features with auxiliary metadata such as age and gender through cross-modal attention, enabling binary classification of image sets. The framework achieves high accuracy with reduced computational parameters, making it suitable for deployment on resource-constrained devices. The primary utility lies in providing a portable, efficient tool for image classification that can support healthcare applications in rural and low-resource environments.
Disclaimer: Curated by HT Syndication.
