Estimating Scalp Moisture in a Hat Using Wearable Sensors.

Area of Science:

• Biomedical Engineering
• Wearable Technology
• Machine Learning Applications

Background:

• Scalp moisture content significantly impacts hair quality, with dryness leading to hair loss and dandruff.
• Continuous monitoring of scalp moisture is crucial for maintaining hair health.

Purpose of the Study:

• To develop a cost-effective, hat-shaped device with wearable sensors for continuous scalp moisture estimation.
• To evaluate the efficacy of machine learning models in predicting scalp moisture using sensor data.

Main Methods:

• Development of a hat-shaped device integrated with wearable sensors for data collection.
• Implementation and comparison of four machine learning models (two non-time-series, two time-series).
• Data collection in a controlled environment for model training and validation.

Main Results:

• Inter-subject evaluation using Support Vector Machine (SVM) achieved a Mean Absolute Error (MAE) of 8.50.
• Intra-subject evaluation using Random Forest (RF) demonstrated an average MAE of 3.29 across all subjects.
• The developed device provides a viable alternative to expensive moisture meters and professional scalp analyzers.

Conclusions:

• The hat-shaped wearable device effectively estimates scalp moisture content using machine learning.
• This technology offers an accessible and affordable solution for individuals to monitor their scalp health.
• Continuous monitoring can aid in the early detection and prevention of scalp-related issues.