Comfort level classification during patients transport.

Area of Science:

• Biomedical Engineering
• Transportation Safety
• Human-Computer Interaction

Background:

• Passenger comfort is influenced by various factors, with patient comfort being uniquely impacted by their health status.
• Existing transport monitoring systems often overlook the specific needs and sensitivities of patients.

Purpose of the Study:

• To develop and evaluate a system for monitoring patient transport conditions and classifying comfort levels.
• To integrate patient-specific parameters into comfort assessment during medical transport.

Main Methods:

• An Android application on a smartphone within an Emergency Medical Service (EMS) vehicle collected GPS data and 10 transport parameters.
• Three machine learning classifiers were employed to categorize transport conditions.
• Comfort level classifications were adjusted based on patient medical condition, age, and gender.

Main Results:

• A modified Support Vector Machine (SVM) classifier achieved the highest classification precision at 90.8%.
• A predictive model for patient sensitivity to transport vibration, considering medical condition, was developed.
• The system demonstrated effective classification of patient transport comfort.

Conclusions:

• The developed Android application is mobile, user-friendly, and suitable for patient transport monitoring.
• SVM and Naive Bayes classifiers yielded satisfactory results, while KNN was less effective.
• The system's ability to incorporate both transport conditions and patient-specific factors makes it valuable for optimizing patient comfort during transit.