Improving Diabetes Mellitus Diagnosis Performance Using DBSCAN Case-Based Reasoning and Clustering Approaches

Diabetes Mellitus (DM) is a chronic disease whose prevalence continues to increase throughout the world, including in Indonesia. The limited number of medical personnel and the large amount of health data that must be managed require technology that can help the diagnosis process faster and more accurately. In this study, a Case-Based Reasoning (CBR) system was developed which was optimized with the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) method to improve the performance of Diabetes Mellitus diagnosis. CBR works by utilizing previous cases to solve new cases, while DBSCAN is used to group large and irregular medical data. This grouping process reduces the complexity of searching for relevant cases, thereby increasing the efficiency of diagnosis. The test results show that the proposed system is able to increase the accuracy of diagnosis by up to 93%, with a shorter case encounter time compared to other methods. The accuracy value was obtained from the results of the diagnosis test using case data of 30 diabetes mellitus patients. The innovation in this study is the development of a CBR-DBSCAN-based diagnostic system in the health system to support medical personnel in providing a fast and accurate diagnosis.