TriHG-XAI: A Human-Grounded Framework for Triangulated Explainability Evaluation in Ordinal Clock Drawing Test Quality Prediction
Diagnostics · 11.11.2026 · 16 (22) · 3735
DOI: 10.3390/diagnostics16223735 ↗Kurz gefasst
Fachartikel in der Zeitschrift Diagnostics, online erschienen am 11.11.2026. Verfasst von Ayat Elmyerigiy und 2 weiteren Autor:innen. Die Arbeit nennt 55 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: KAU Endowment (WAQF), King Abdulaziz University, Jeddah, Saudi Arabia.
Abstract (Originalsprache)
Background/Objectives: Deep learning models can score the Clock Drawing Test (CDT) accurately, but their explanations are usually judged by looking at saliency maps. We asked whether complementary quantitative explainability criteria, evaluated against independent human annotations, give convergent evidence that a CDT model has learned interpretable drawing abnormalities.
Methods
An image-only ResNet50 was trained to predict five-level CDT scores from the National Health and Aging Trends Study (7708 observations; participant-level split). In 200 separate, participant-unique drawings annotated for seven abnormalities, we evaluated Grad-CAM attribution alignment, matched-ink perturbation sensitivity, and linear probing of frozen representations. Sensitivity analyses covered region-of-interest definitions, data quality, and adjustment for CDT score.
Results
On 1181 test observations, the quadratic weighted kappa was 0.740 and accuracy 0.678. Unadjusted analyses suggested abnormality-specific effects, including reduced hand-region attribution for missing or incorrect hands (q < 0.001) and significant perturbation sensitivity for three abnormalities. However, perturbation effects were as large in drawings without the abnormality. Within score levels, neither attribution nor perturbation sensitivity distinguished drawings with and without abnormalities. Within-score linear decodability remained for five abnormalities (AUROC 0.64–0.93), but not for uneven number spacing or same-length hands. Inter-rater agreement varied (κ 0.27–0.85).
Conclusions
In this model and dataset, attribution and perturbation sensitivity mainly reflected overall drawing quality and the presence of intact hands and numerals. Abnormality-specific information was detectable only in the learned representation. Feature-level explainability evaluations should control for the prediction target, and saliency maps alone do not show that a model has learned interpretable abnormalities.
Förderung
- KAU Endowment (WAQF), King Abdulaziz University, Jeddah, Saudi Arabia
Zitieren
APA
Elmyerigiy, A., Jaha, E. S., & Bafail, A. (2026). TriHG-XAI: A Human-Grounded Framework for Triangulated Explainability Evaluation in Ordinal Clock Drawing Test Quality Prediction. Diagnostics, 16(22), 3735. https://doi.org/10.3390/diagnostics16223735
Vancouver
Elmyerigiy A, Jaha ES, Bafail A. TriHG-XAI: A Human-Grounded Framework for Triangulated Explainability Evaluation in Ordinal Clock Drawing Test Quality Prediction. Diagnostics. 2026;16(22):3735. doi:10.3390/diagnostics16223735
BibTeX
@article{Elmyerigiy2026,
author = {Elmyerigiy, Ayat and Jaha, Emad Sami and Bafail, Alaa},
title = {TriHG-XAI: A Human-Grounded Framework for Triangulated Explainability Evaluation in Ordinal Clock Drawing Test Quality Prediction},
journal = {Diagnostics},
year = {2026},
volume = {16},
number = {22},
pages = {3735},
doi = {10.3390/diagnostics16223735}
}
EUVIMED