Video transformers quantify eye tics in adolescents with Tourette syndrome
Source check: Source checked. Not independent scientific confirmation.
Finding: A transformer classified eye-tic video segments from 11 adolescents with Tourette syndrome. Performance fell from an AUC of about 0.90 under random splitting to about 0.75 when patients were separated between training and validation. Validation data were monitored during optimization, so this is limited internal diagnostic measurement evidence rather than an untouched external test, a Tourette diagnosis or demonstrated clinical benefit.
Technology: Transformer; computer-vision facial landmarks
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
- AI contribution
- A shallow transformer learns temporal sequences of facial landmarks to identify manually annotated eye-tic events.
- Test and result
- Manually labeled eye-tic segments and patient-separated validation, alongside a random-split comparison.
- Limitation / still unknown
- External patient testing, broader tic recognition and clinical utility remain unestablished; monitored validation is not a pristine test set.
Evidence sources and annotation context
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC10895867/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 21 Nov 2026. Review scope: Primary relevant Methods/results/limitations and exact front/canonical identity; current primary PubMed linked correction/retraction metadata checked..
No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.
Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.