Deep-Learning-Based ADHD Classification Using Children's Skeleton Data Acquired through the ADHD Screening Game
Source check: Source checked. Not independent scientific confirmation.
Finding: A bidirectional LSTM with weighted loss classified clinician-labeled ADHD, ADHD-risk and normal groups from game-acquired skeleton sequences in 596 children; best reported accuracy was 97.82% under leave-one-person-out validation. The source records ADHD-risk cases classified as normal and frames the system as screening support, not diagnosis. Methods §3.4: leave-one-person-out; labels incorporate clinician review and CBCL/K-ARS, but are not independent diagnoses; limitations in §5.
Technology: Methods §3.4: GRU/RNN/LSTM sequence classifier, bidirectional layer and weighted cross-entropy on 18-joint skeleton sequences.
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
- AI contribution
- Methods §3.4: GRU/RNN/LSTM sequence classifier, bidirectional layer and weighted cross-entropy on 18-joint skeleton sequences.
- Test and result
- Methods §3.3, Results §4: 596 children and three clinician-labeled groups; 97.82% LSTM and ADHD-risk false-normal errors.
- Limitation / still unknown
- Methods §3.4: leave-one-person-out; labels incorporate clinician review and CBCL/K-ARS, but are not independent diagnoses; limitations in §5. The source does not establish benefits beyond the reported evaluation.
Evidence sources and annotation context
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
Evidence sources: https://www.mdpi.com/1424-8220/23/1/246 · https://europepmc.org/article/MED/36616844
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: other. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 22 Nov 2026. Review scope: Primary result-bearing material for the stated method, result and evaluation, with bibliographic version information..
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.