A Low-Cost Wearable Device to Estimate Body Temperature Based on Wrist Temperature

Area of Science:

• Biomedical Engineering
• Health Informatics
• Wearable Technology

Background:

• The COVID-19 pandemic highlighted the need for remote vital sign monitoring.
• Blood oxygen, heart rate, and body temperature are critical for disease management.
• Non-contact, continuous monitoring of these vitals is essential for timely medical care.

Purpose of the Study:

• To propose a low-cost wearable device for remote monitoring of vital signs.
• To focus on body temperature measurement as a key health indicator.
• To develop methods for estimating core forehead temperature from wrist temperature data.

Main Methods:

• A low-cost wearable device with non-contact sensors was developed.
• Data on wrist and forehead temperature were collected from 30 volunteers (2130 samples).
• Statistical (least squares) and machine learning (linear regression) models were used to estimate forehead temperature from wrist data.

Main Results:

• Both statistical and machine learning models showed significant fits for temperature estimation.
• A third-degree polynomial model demonstrated superior performance in both approaches.
• The third-degree polynomial model achieved R² values of 0.9769 (statistical) and 0.9791 (machine learning).

Conclusions:

• The developed wearable device offers a viable non-invasive method for continuous temperature monitoring.
• Statistical and machine learning approaches can effectively estimate core body temperature from wrist measurements.
• The third-degree polynomial model shows high accuracy in predicting forehead temperature, aiding remote patient care.