Physiological Status Prediction Based on a Novel Hybrid Intelligent Scheme

Area of Science:

• Biomedical Engineering
• Health Informatics
• Machine Learning in Healthcare

Background:

• Physiological status is critical for clinical diagnosis, but temporal data is dynamic and challenging to manage.
• Obtaining complete historical physiological data is often difficult, hindering accurate diagnosis.

Purpose of the Study:

• To develop a hybrid intelligent scheme for accurate physiological status prediction.
• To provide a reliable reference for clinical diagnosis using temporal physiological data.

Main Methods:

• Extracted attribute information from nonlinear dynamic physiological signals.
• Selected optimal features using conditional relevance mutual information.
• Employed particle swarm optimization-support vector machine (PSO-SVM) for classification.

Main Results:

• The hybrid intelligent scheme demonstrated superior performance in predicting sleep status on the Sleep Heart Health Study dataset.
• The proposed method outperformed conventional machine learning classification techniques.

Conclusions:

• The developed hybrid intelligent scheme offers an effective solution for physiological status prediction.
• This approach can enhance clinical decision-making by providing accurate patient status insights.