ASD-Bench: A Four-Axis Comprehensive Benchmark of AI Models for Autism Spectrum Disorder
Provides a comprehensive benchmark for automated ASD screening, highlighting age-specific diagnostic patterns and the need for multi-axis evaluation, but results are proof-of-concept on questionnaire-derived labels.
ASD-Bench evaluates 17 ML/DL models on ASD screening across three age cohorts using a curated AQ-10 dataset, finding that adult classification achieves perfect F1/AUC for most models, while adolescents are harder (F1 ceiling 0.837). Feature importance shifts across cohorts, and accuracy-calibration dissociation (e.g., AdaBoost F1=1.000 but ECE=0.302) shows single-metric evaluation is insufficient.
Automated ASD screening tools remain limited by single-architecture evaluations, axis-restricted assessment, and near-exclusive focus on adult cohorts, obscuring age-specific diagnostic patterns critical for early intervention. We introduce ASD-Bench, a systematic tabular benchmark evaluating ML, deep learning, and foundation model configurations across three age cohorts (children 1-11 yr, adolescents 12-16 yr, adults 17-64 yr) on four axes: predictive performance, calibration, interpretability, and adversarial robustness. Applied to a curated v3 dataset of 4,068 AQ-10 records, our benchmark spans classical models (XGBoost, AdaBoost, Random Forest, Logistic Regression), neural networks (MLP), deep tabular transformers (TabNet, TabTransformer, FT-Transformer), and TabPFN v2. We introduce the Heuristic Aggregate Penalty (HAP): a cost-sensitive metric penalising false negatives more heavily and incorporating cross-validation variance for deployment stability. Adult classification yields high performance (10/17 models achieve perfect F1 and AUC), while adolescents present a harder task (F1 ceiling 0.837 vs. 0.915 for children). Feature hierarchies shift across cohorts: A9 (social motivation) dominates for children, A5 (pattern recognition) leads for adolescents, and adults exhibit a flatter importance profile consistent with developmental social masking. Accuracy and calibration are dissociated: AdaBoost achieves F1=1.000 on adults with ECE=0.302, confirming single-metric evaluation is insufficient for clinical AI. Cohort-specific deployment recommendations are provided. All findings should be interpreted as proof-of-concept evidence on questionnaire-derived labels rather than clinically validated diagnostic performance.