DEVELOPMENT OF A WEARABLE DEVICE FOR ANXIETY MONITORING USING GALVANIC SKIN RESPONSE AND PHOTOPLETHYSMOGRAPHY
Mental health, particularly anxiety disorders, has gained increasing global attention, with approximately 970 million people affected in 2019, and the numbers rising during the COVID- 19 pandemic. Anxiety disorders, which impact the psychological, emotional, and physiological aspects of indiv...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Project |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/86056 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | Mental health, particularly anxiety disorders, has gained increasing global attention, with
approximately 970 million people affected in 2019, and the numbers rising during the COVID-
19 pandemic. Anxiety disorders, which impact the psychological, emotional, and physiological
aspects of individuals, can disrupt daily life and elevate the risk of serious health issues if left
untreated. Early detection is crucial, and the advancement of wearable device technology
enables the monitoring of physiological signals, such as heart rate and skin response, to detect
signs of anxiety. Therefore, this research aims to design a wearable device capable of detecting
anxiety based on data from PPG and GSR sensors, helping individuals monitor their mental
health and providing early warnings when high levels of anxiety are detected. The developed
wearable device system processes data directly on the ESP-WROOM-32 microcontroller, from
data acquisition to anxiety prediction. Through acquisition and pre-processing, this device can
extract key features such as Heart Rate Variability (HRV) and Skin Conductance Response
(SCR) related to anxiety. Testing showed that the RR Interval data obtained using the PPG
sensor demonstrated a Mean Absolute Error (MAE) of 9,03 ms compared to RR interval data
from Biopac. Meanwhile, the GSR data exhibited similar trends between the utilized sensor
and Biopac. Additionally, a model trained using the HRV dataset with a neural network
approach achieved 92% accuracy. The testing results also indicated that the performance of
the converted model running on the microcontroller was comparable to the model executed in
Python. However, the microcontroller's memory limitations necessitate algorithm
simplification to maintain performance and accuracy. Further studies are needed to address
these challenges and enhance the accuracy of wearable devices in anxiety detection. |
|---|