This research explores the implementation of IoT and machine learning models for heart disease prediction and diagnosis in the healthcare domain. By employing various IoT sensors, including ECG, oximeter, and heart rate monitoring devices, patient health is continuously monitored, and data is collected. This data is then processed using machine learning techniques to develop predictive models capable of detecting heart disease at an early stage. Two dissimilar machine learning methods are employed: linear regression and artificial neural networks (ANNs). The linear regression model utilizes historical data to generate an equation, achieving an accuracy of 95.76% in predicting heart disease. In contrast, the ANN model, incorporating a feedforward algorithm with 10 neurons, is trained using the TRAINLM training function and the LEARNGDM adaptation learning function, achieving an impressive accuracy of 100%. The results demonstrate the superiority of the ANN model in terms of accuracy, highlighting its ability to capture complex nonlinear relationships within the data. By accurately predicting heart disease, these models serve as valuable tools for healthcare professionals to intervene promptly and provide appropriate treatments, leading to improved patient outcomes and reduced healthcare costs. Moreover, the research establishes normal ranges for sensor readings, such as ECG, oximeter, and heart rate, enabling the identification of normal and abnormal conditions. This facilitates the timely identification of individuals at risk and the initiation of necessary actions. Overall, this research showcases the potential of IoT and machine learning in revolutionizing healthcare through early detection and diagnosis of heart disease, fostering proactive and personalized healthcare practices, and ultimately improving patient care and optimizing healthcare resource utilization. Future research can focus on expanding datasets, incorporating additional sensors, and further refining the accuracy and reliability of heart disease prediction models for real-time clinical applications.