Facial biomarkers
Expression variability
Mean temporal standard deviation of the expression blendshapes.
- Key
expressivityVariability- Unit
- 0–1
- RDoC domain
- Positive valence systems
- For the sample
- 0.0264 0–1
Method
Standard deviation over frames, averaged across expression blendshapes.
Quality gate
At least 10 s of frames with a detected face. Below the gate the value is null and the reason is listed in the response's quality flags.
How the engine reads it
In Positive valence systems, with lower values than the person's own baseline read as concern, at full weight. Flatter, less changing expression.
Published associations
| Condition | In the condition | Evidence |
|---|---|---|
| Depression | ↓lower | replicatedGaebel W, Wölwer W; Davies H, Wolz I, Leppanen J, et al |
| Schizophrenia spectrum | ↓lower | replicatedGaebel W, Wölwer W; Tron T, Peled A, Grinsphoon A, Weinshall D |
Group-level findings from the literature, never a diagnosis. Compare a person with themselves over time.
References
- 1.Google AI Edge. MediaPipe Face Landmarker: 478 landmarks and 52 blendshape scores. developers.google.com/edge/mediapipe/solutions/vision/face_landmarker
- 2.Gaebel W, Wölwer W. Facial expressivity in the course of schizophrenia and depression. Eur Arch Psychiatry Clin Neurosci 2004;254:335-342. doi.org/10.1007/s00406-004-0510-5
- 3.Davies H, Wolz I, Leppanen J, et al. Facial expression to emotional stimuli in non-psychotic disorders: a systematic review and meta-analysis. Neurosci Biobehav Rev 2016;64:252-271. doi.org/10.1016/j.neubiorev.2016.02.015
- 4.Tron T, Peled A, Grinsphoon A, Weinshall D. Facial expressions and flat affect in schizophrenia, automatic analysis from depth camera data. IEEE BHI 2016:220-223. doi.org/10.1109/BHI.2016.7455874
Measure it on your own data.
Sandbox today. Production under agreement.
Or make a first request with the sandbox key.