MUMBAI, India, July 24 -- Intellectual Property India has published a patent application (202641085998 A) filed by Chennai Institute Of Technology on July 14, 2026, for Non-Invasive Identification Of Blood Group Determination Method And System Using Fingerprint Biometrics And Deep Learning.
Inventors include Dr. S. Radhika; Mrs. M. Harshanya; Mr. S. Vignesh Kumar; and Mrs. G. Latha.
The application for the patent was published on July 17, 2026, under issue no. 29/2026.
Abstract: Identification of a blood group is an important task, especially during medical emergency, such as surgical procedure, transfusion, where it is required to keep blood ready for which blood group is essential. The conventional methods for identifying through laboratory procedure, the blood group is identified. This type of procedure may not hold good when the requirement is in a remote place or when there is less resource available or when the time is very short. During these situations, it is essential to have a non- invasive method for the identification of blood group. So the developed system addresses these problems where the fingerprint of the individual is collected and it is fed to the deep learning technique. It is prior identified that each individual has a unique fingerprint pattern and there is a correlation existing between the fingerprint and the blood group. So, using this correlation extraction of features from the fingerprint of the individual is done. And then using transfer learning with predefined ImageNet weights, the model efficiency is improved and hence the classification performance is also improved further. Using 70% of training and 30% of testing datasets, and by evaluating through various performance metrics, the validation is done. Further, the system when deployed on the mobile phones or in health centres, is found to provide highly accurate results and also efficient. The chief advantages claimed by the system are good reliability, lesser cost and finally, non-invasive thus contributing to a faster and accurate diagnosis solutions.
Disclaimer: Curated by HT Syndication.
