Kernel random forest with black hole optimization for heart diseases prediction using data fusion.

Area of Science:

• Biomedical Engineering
• Artificial Intelligence
• Health Informatics

Background:

• The Internet of Medical Things (IoMT) enables remote patient monitoring via wearable sensors.
• Heart disease diagnosis relies on multi-sensor signal fusion, but existing methods face accuracy and efficiency challenges.

Purpose of the Study:

• To develop an efficient algorithm for fusing multi-sensor signals and classifying medical data for accurate heart disease prediction.
• To address limitations in accuracy, time consumption, and efficiency in current diagnostic approaches.

Main Methods:

• Proposed a hybrid technique combining Kernel Random Forest with Black Hole Optimization (KRF-BHO) for sensor data fusion.
• Utilized XGBoost classifier for analyzing echocardiogram images and medical signal data.
• Evaluated the model on multi-sensor data fusion and the Cleveland heart disease dataset.

Main Results:

• Achieved 95.89% accuracy in the testing phase using the multi-sensor data fusion dataset.
• Attained 96.21% accuracy in the testing phase with the Cleveland Dataset.
• Demonstrated high performance in both training and testing phases for heart disease prediction.

Conclusions:

• The proposed KRF-BHO and XGBoost hybrid model offers an efficient and accurate solution for heart disease prediction.
• This approach effectively overcomes the limitations of existing methods in terms of accuracy and efficiency.
• Highlights the potential of IoMT and advanced machine learning for improving cardiovascular diagnostics.