Automating Heartbeat Disease Detection Using Advanced Machine Learning Algorithm

Heart diseases remain a top contributor to death rates globally, posing significant challenges. Machine learning is becoming more valuable in medical diagnostics by providing novel methods to address various health issues. This research applies Convolutional Neural Network (CNN) model to automate the detection of heartbeat diseases (including Artifact, Murmur, Extrastole, and Extrahls) using 1,441 heartbeat sound records as inputs. By carefully pre-processing heartbeat sounds and training the CNN model, we successfully obtained refined feature extraction, which is crucial for precise classification. A standard evaluation criterion is applied for a comprehensive evaluation. The model's performance examination demonstrated its ability in identifying heartbeats linked to different heart conditions, where the training, and validation accuracies reached 98 %, and 92 %, respectively. The model demonstrated considerable promise with a peak precision rate of 100% for identifying Extrastole disease. These results are excellent considering the complexity associated with detecting heartbeat diseases.