Enhancing Hematological Diagnostics: A Machine Learning Framework Featuring Michael’s Classifier for Beta Thalassemia

Beta thalassemia is a genetic disorder that causes significant health and economic challenges, particularly in regions like the Mediterranean, the Middle East, and South Asia, due to the lifelong medical care required. Although advances in machine learning and deep learning have improved disease diagnosis, current models often struggle with accurately diagnosing beta thalassemia due to the disease’s complexity and the variability in the quality of medical data. Despite numerous studies applying artificial intelligence to genetic diseases, the use of critical data, such as gene sequences, medical records, and diagnoses, remains limited, which impedes the accuracy and generalizability of these models across different populations. This study aims to develop a deep learning model that integrates both genetic data and clinical images to provide a more precise and meaningful diagnosis of beta thalassemia, addressing the limitations of existing diagnostic methods. We employed a deep learning architecture that combines convolutional neural networks (CNN) for image analysis with long short-term memory (LSTM) networks for data processing. While more rigorous training (with a training accuracy of 0.75) raised concerns about overfitting due to the disparity in performance between the training (0.75), validation (1.00), and testing (1.00) stages, the proposed model holds promise for significantly enhancing diagnostic accuracy and efficiency. This is particularly beneficial in healthcare settings with limited resources, enabling more timely treatment and better overall patient outcomes.