Blood Group Identification by Analyzing Fingerprints using ResNet-50

In the domain of medical diagnostics, identification of blood groups accurately and efficiently is essential for various clinical procedures, including transfusions and organ transplants. Traditional invasive procedures involve taking a blood sample from individuals, which can be uncomfortable and may cause infections. Additionally, testing blood samples in the laboratory is time-consuming. The present approach proposes a non-invasive method for blood group identification using deep learning and fingerprint images. Fingerprints are unique patterns found on our fingertips that help in identifying individuals. By using these unique and unchangeable features of fingerprints, every individual’s blood group can be identified. The proposed approach leverages deep learning model ResNet-50, which is well known for its significance in image classification, to predict blood groups efficiently. The model achieved an accuracy of 88%, confirming its effectiveness in blood group identification. This approach addresses a critical need in the medical community, providing a seamless and technologically advanced solution for blood group identification