Towards depression monitoring and prevention in older populations using smart wearables: Quantitative Findings
- Publisher:
- IEEE
- Publication Type:
- Conference Proceeding
- Citation:
- SeGAH 2022 - 2022 IEEE 10th International Conference on Serious Games and Applications for Health, 2022, 00, pp. 1-8
- Issue Date:
- 2022-01-01
Closed Access
| Filename | Description | Size | |||
|---|---|---|---|---|---|
| Towards_depression_monitoring_and_prevention_in_older_populations_using_smart_wearables_Quantitative_Findings.pdf | 678.36 kB |
Copyright Clearance Process
- Recently Added
- In Progress
- Closed Access
This item is closed access and not available.
Depression has become a growing concern over the recent years. Since the start of the COVID-19 pandemic, depression among all age groups has increased significantly. As mental health is often stigmatized among older aged people, it is less openly discussed or treated. We propose a mental health monitoring approach that limits explicit user interaction, using Fitbit smartwatch data to determine depressive tendencies in older-aged people. We analysed physiological user data extracted from a Fitbit Alta HR device and use this data to train a machine learning model to detect depressive tendencies. While this is not a diagnostic tool, the aim is to identify physiological signs early on and direct the user toward professional medical guidance and treatment. We trained 19 predictive models on our dataset, the gradient boosting regressor outperformed all other models. The best performing model achieved at R-square of 0.32 although most models were poorly performing. Due to the limited sample size, there is a risk of model overfitting. Although these preliminary results are promising for one model, they would need to be replicated in a larger sample of older people, who exhibit a wider range of depressive tendencies.
Please use this identifier to cite or link to this item: