Mental Health Risk
The Mental Health Risk metric offers a comprehensive, holistic approach to evaluating an individual’s potential risk of developing a mental health condition over time. Such conditions are characterised by significant disturbances in cognition, emotional regulation, or behaviour.
This metric is a clinically validated benchmark that aligns with standard assessments like the PHQ-9, GAD-7, and CSAI-2 questionnaires, focusing on physiological markers of stress and anxiety.
Name | Unit | programmatic name |
|---|---|---|
Mental Health Risk | Not applicable | mental_health_risk |
IntelliProve’s processing engine uses machine learning informed by HRV metrics, facial micro-expressions, and context-based reasoning to assess Mental Health Risk. By comparing these factors over time, the metric provides a risk level of low, medium, or high relative to the user’s baseline.
When is the first reading available? To receive an initial reading, users must complete at least two Face Scans per week for three consecutive weeks. The first score will be available after the second Face Scan in the third week.
Example timeline with Face Scans:
- Week 1: Monday & Wednesday - No MHR available yet.
- Week 2: Tuesday & Friday - No MHR available yet.
- Week 3: Monday & Tuesday - First MHR score displayed on Wednesday
Participants were asked to fill in a combination of questionnaires to assess the mental health status of the individual in terms of depression (PHQ-9) and anxiety (GAD-7). Based on the outcome of these questionnaires, every participant was classified according to a certain mental health risk profile (low, medium or high):
- Low: GAD-7<5 AND PHQ-9<5
- Medium: 5≤GAD-7≤9 AND/OR 5≤PHQ-9≤9
- High: GAD-7>9 AND/OR PHQ-9>9
It can be concluded that in 100% of the cases participants with a high mental health risk will also be labeled as ‘high risk’ through the IntelliProve processing engine (true positive cases). Although the discriminative power of IntelliProve between medium and low risk profiles is substantially lower, it can be concluded that 89.1% of the combined medium/low cases will also be labeled as medium/low through IntelliProve.
The Mental Health Risk metric is particularly valuable in preventive care by identifying negative trends early. While one bad day doesn’t imply a high Mental Health Risk, a continuous series of negative days or a prolonged decline may signal potential issues, such as burnout.
By providing a risk score, IntelliProve raises awareness about potential concerns, helping users recognize early signs and encouraging timely intervention. In cases of high risk, users could be advised to consult with a mental health professional for further support. This objective biomarker complements subjective self-reporting, offering a fuller picture of the user’s mental health status.
Providing users with a risk score will raise awareness about potential concerns, such as moving towards a long-term absence. When a high risk is detected, users could be recommended to schedule a call with a therapist or coach for further support and guidance.
This risk assessment serves as an ideal, more objective metric, complementing the subjective user reports.
Mental Health Risk is returned as a value between 1 and 3 and can be requested as a Widget or via the Rest API.
Value | Definition | Example Color Indicator | Example User Text |
|---|---|---|---|
1 | Low mental health risk | Green | You have a lower risk of mental health conditions. |
2 | Medium mental health risk | Yellow | You have a medium risk of mental health conditions. |
3 | High mental health risk | Red | You have a higher risk of mental health conditions. We recommend that you pursue a further discussion of your situation with a physician or mental health provider. |
- Kemp AH, Quintana DS, Gray MA, Felmingham KL, Brown K, Gatt JM. 2010. Impact of depression and antidepressant treatment on heart rate variability: a review and meta-analysis. Biol Psychiatry. 67(11):1067–1074.
- Wang X, Wang Y, Zhou M, Li B, Liu X, Zhu T. Identifying Psychological Symptoms Based on Facial Movements. Front Psychiatry. 2020 Dec 15;11:607890.
- Sharma, D., Singh, J., Sehra, S. S., & Sehra, S. K. (2024). Demystifying Mental Health by Decoding Facial Action Unit Sequences. Big Data and Cognitive Computing, 8(7), 78.
- Licht CM, de Geus EJ, Zitman FG, Hoogendijk WJ, van Dyck R, Penninx BW. 2008. Association between major depressive disorder and heart rate variability in the Netherlands Study of Depression and Anxiety (NESDA). Arch Gen Psychiatry. 65(12):1358–1367.
- Thayer JF, Åhs F, Fredrikson M, Sollers JJ 3rd, Wager TD. 2012. A meta-analysis of heart rate variability and neuroimaging studies: implications for heart rate variability as a marker of stress and health. Neurosci Biobehav Rev. 36(2):747–756.