This project applies anomaly detection techniques, specifically DBSCAN and Local Outlier Factor (LOF), to analyze a hypothyroidism dataset containing 7,200 entries and 21 attributes. Through comprehensive data preprocessing and parameter optimization, including the use of Gower distance and visualizations like t-SNE, the analysis successfully identified significant anomalies, highlighting subtle disease patterns potentially overlooked by traditional methods. The combined DBSCAN and LOF approach proved particularly effective, achieving an adjusted Rand index of 0.6322, indicating strong consistency between these methods in anomaly detection.