Explainable AI for Early Detection and Classification of Childhood Leukemia Using Multi-Modal Medical Data

Childhood leukemia is one of the most common forms of cancer in children, making early detection and accurate diagnosis crucial for improving patient outcomes. Traditional methods of diagnosing leukemia, including manual analysis of medical images and blood tests, are often time-consuming and prone to human error. Recent advancements in artificial intelligence (AI) have shown promise in improving diagnostic accuracy and speed. This paper presents an explainable AI (XAI) approach for the early detection and classification of childhood leukemia using multi-modal medical data, including blood test results, imaging data, and clinical history. Our proposed system integrates deep learning models for image classification with machine learning algorithms for structured data analysis, resulting in a multi-modal framework capable of not only detecting leukemia at an early stage but also providing interpretable predictions. This paper outlines the methodology, system architecture, and experimental setup used to build the model, as well as the results obtained from evaluation using publicly available medical datasets. We demonstrate that the integration of multi-modal data improves classification accuracy compared to single-modal systems and that the explainability of the AI model helps healthcare professionals interpret the results, thus increasing trust in AI-driven decision-making systems. Finally, we discuss the future implications of this approach and potential areas for further development.