Enhancing Medical Big Data Analytics: A Hadoop and FP-Growth Algorithm Approach for Cloud Computing

Effective mining of relationships within massive medical datasets can profoundly enhance clinical decision-making and healthcare outcomes. However, traditional data mining techniques falter in extracting actionable associations from large-scale medical data. This research optimizes the Frequent Pattern Growth algorithm and incorporates it into a Hadoop framework for scalable medical data analytics. Empirical evaluations on real-world patient diagnosis records demonstrate the proposed approach's computational and learning efficiency. For instance, with the Break-Cancer database, the optimized algorithm requires just 0.04 seconds at 0.22 minimum support, significantly faster than existing methods. Experiments on diagnostics data generate 267 informative association rules at 0.31 support - markedly higher than 71, 126 and 233 rules produced by other comparative techniques. By enabling rapid discovery of data-driven health insights, the enhanced medical data mining framework provides a valuable decision-support system for better clinical practice. Ongoing explorations focus on further optimizations for automated disease prediction and treatment recommendations to continuously augment data-to-diagnosis applicability.