Detection of Intermittent Claudication from Smartphone Inertial Data in Community Walks Using Machine Learning...

Area of Science:

• Biomedical Engineering
• Cardiovascular Research
• Digital Health

Background:

• Peripheral arterial disease (PAD) obstructs lower limb arteries, causing significant pain.
• Claudication onset is a critical indicator of PAD progression.
• Current detection methods can be invasive or lack real-time monitoring.

Purpose of the Study:

• To develop an automated method for detecting the onset of claudication.
• To leverage smartphone inertial sensors for PAD symptom monitoring.
• To evaluate machine learning models for accurate claudication detection.

Main Methods:

• Collected gait data from 40 PAD patients using smartphones over a 6-minute walk test.
• Performed data analysis on two distinct datasets derived from patient recordings.
• Compared various machine learning models to identify the onset of claudication.

Main Results:

• The Extreme Gradient Boosting model achieved the highest accuracy of 92.25% in detecting claudication onset.
• Inertial sensor data from smartphones proved effective for symptom identification.
• The study demonstrated the feasibility of remote, automated PAD symptom monitoring.

Conclusions:

• Smartphone-based inertial sensing combined with machine learning offers a promising, non-invasive approach for early claudication detection.
• This technology can facilitate timely intervention and improve management of peripheral arterial disease.
• Further research could integrate this method into wearable devices for continuous patient monitoring.