Improving Neurocognitive Testing Using Computational Psychiatry-A Systematic Review for ADHD

Computational models, in conjunction with (neuro)cognitive tests, are increasingly used to understand the cognitive characteristics of participants with attention-deficit/hyperactivity disorder (ADHD). We reviewed 50 studies from a broad range of cognitive tests for ADHD to synthesize findings and to summarize the new insights provided by three commonly applied computational models (i.e., diffusion decision models, absolute accumulator models, ex-Gaussian distribution models). Four areas are discussed to improve the utility of (neuro)cognitive testing for ADHD: (a) the requirements for appropriate application of the computational models; (b) the consideration of sample characteristics and neurophysiological measures; (c) the integration of findings from cognitive psychology into the literature of cognitive testing to reconcile mixed evidence; and (d) future directions for the study of ADHD endophenotypes. We illustrate how computational models refine our understanding of cognitive concepts (slow processing speed, inhibition failures) presumed to characterize ADHD. We also show that considering sample characteristics and integrating findings from computational models and neurophysiological measures provide evidence for ADHD endophenotype-specific cognitive characteristics. However, studying the cognitive characteristics of ADHD endophenotypes often lies beyond the scope of existing research for three reasons: some cognitive tests lack sensitivity to detect clinical characteristics; analysis methods do not allow the study of subtle cognitive differences; and the precategorization of participants restricts the study of symptom severity on a continuous spectrum. We provide recommendations for cognitive testing. computational modeling, and integrating electrophysiological measures to produce more valuable tools in research and clinical practice (above and beyond the research domain of ADHD).