Heart disease is one of the major causes of death all over the world, and there is a need to find an early intervention and prevention therapy. This paper offers an original model that combines the Internet of Medical Things (IoMT) with a sophisticated hybrid AI model to make correct and timely predictions of heart disease. It is based on IoMT architecture that relies on wearable and implantable devices to gather multimodal data, such as ECG, heart rate, blood pressure, oxygen saturation, and environmental indicators, such as temperature and air quality index. These streams are securely sent to edge and cloud systems to be preprocessed, and noise filtering and feature extraction are carried out to improve the quality of data. The proposed hybrid AI architecture is a combination of Convolutional Neural Networks (CNNs) to extract spatial features and Transformer-based architecture to analyze sequential data. The model syndicates spatial and temporal features, which results in a high performance of predicting heart disease risk. The model was assessed on publicly available datasets, like the PhysioNet and the Cleveland Heart Disease Dataset, also it was shown to be highly accurate (94.3%), precise (93.1%), and dependable in a variety of states. The proposed framework demonstrated promising performance in simulated IoMT healthcare environments and showed potential for supporting real-time cardiovascular risk assessment. This paper underscores the potential of the IoMT and hybrid AI approaches to transform cardiovascular healthcare to provide scalable and effective cardiovascular care by predicting heart disease early and managing it proactively. Further development of the framework will be done by increasing the diversity of the data set and hold large-scale clinical trials to test and refine the framework.