A previous heart disease diagnosis system relied on Interval Type-2 Fuzzy Logic System (IT2FLS), but it had poor recognition accuracy and training time. This research proposes an efficient heart disease prediction system using modified firefly algorithm based radial basis function with support vector machine (MFA and RBF-SVM). In smart healthcare AI systems, the decision support component acts as a watchdog or adviser to cut down on medical mistakes, which include patient errors. There are two primary versions of this component. The prediction of CVD makes use of a number of ML and DL methods. On the other hand, a fully clinical setting is ideal for these approaches. However, a new need for improved prediction systems that are suited to real-time monitoring and illness prediction has arisen as a result of technological advancements. To minimizes computing cost and boost prediction system performance, the ideal subset of attributes is selected using the PSO algorithm and an attribute reduction approach based on Rough Sets (RS).