A novel approach for chewing detection based on a wearable PPG sensor.

Area of Science:

• Biomedical Engineering
• Wearable Technology
• Health Monitoring

Background:

• Monitoring human eating behavior is crucial for healthy lifestyles and managing conditions like obesity and eating disorders.
• Traditional self-report methods for monitoring eating behavior are often unreliable.
• Wearable sensors, such as acoustic or strain sensors, show promise for automated monitoring.

Purpose of the Study:

• To introduce a novel photoplethysmography (PPG) sensor for chewing detection.
• To evaluate the effectiveness of a processing pipeline with multiple algorithms for analyzing PPG signals during eating.
• To assess the potential of earlobe-based PPG sensing for dietary monitoring.

Main Methods:

• Development of a novel PPG sensor placed on the outer earlobe.
• Implementation of a processing pipeline incorporating two existing and one new chewing detection algorithm.
• Collection of an annotated dataset from 21 individuals, comprising over 10 hours of eating and non-eating activities.

Main Results:

• The novel earlobe PPG sensor successfully captured signals indicative of chewing activity.
• The processing pipeline demonstrated effectiveness in detecting chewing events from the PPG signals.
• Experimental results validated the PPG sensor's capability to support dietary monitoring.

Conclusions:

• A novel earlobe PPG sensor presents a viable and reliable solution for automated chewing detection.
• This technology can significantly enhance the accuracy and objectivity of dietary monitoring.
• Wearable PPG sensors offer a promising avenue for promoting healthy eating habits and managing weight-related health conditions.