Blood Group Prediction Using Fingerprint

Abstract

A crucial part of medical diagnostics is blood group detection, which is typically done using serological techniques. Novel approaches to non-invasive blood group detection have been made possible by recent developments in computer vision and machine learning. The usefulness of Convolutional Neural Networks (CNNs) for blood type identification from fingerprint photographs is investigated in this paper. Biometric fingerprints are distinct identifiers that have the ability to encode biological data, such as blood type. The suggested approach entails gathering a varied dataset of blood group-related fingerprint pictures. Using this information, a CNN model is created and trained to identify patterns and characteristics typical of various blood groups. To find the best accurate and computationally efficient model, a variety of CNN topologies are compared. Metrics including accuracy, precision, recall, and F1-score are used to evaluate the CNN model's performance. According to preliminary findings, the CNN-based strategy can attain impressive accuracy levels, offering a competitive substitute for conventional blood group identification techniques. Enhancing the model's accuracy, growing the dataset, and resolving any potential privacy and ethical issues with the use of biometric data will be the main goals of future study. This work is a groundbreaking step toward the medical diagnostics industry's integration of biometric data with cutting-edge machine learning algorithms.