<![CDATA[Biomedical psychology emphasises psychological and neurocognitive assessment through the integration of biological, neurophysiological, and quantitative behavioural data to support clinical decision-making. However, conventional assessment approaches remain limited by issues of objectivity, scalability, and longitudinal monitoring, prompting the utilisation of artificial intelligence (AI) as a computational tool in clinical and cognitive contexts. This systematic review synthesises the application of AI in biomedical psychology with an explicit focus on assessment functions, rather than intervention or therapy, following the PRISMA 2020 guidelines through a systematic search of four major databases. The included studies cover a variety of clinical and cognitive applications with variations in psychological constructs, data modalities, and AI methods. The synthesis results show that AI is most often used for diagnostic classification, risk screening, and continuous estimation of cognitive functions and dimensional constructs. Differences in assessment objectives between clinical and cognitive domains reveal consistent methodological trade-offs related to model selection, validation strategies, and overfitting risks. As a key contribution, this review presents an assessment-oriented cross-domain synthesis and proposes fit-forpurpose design principles as a conceptual framework for developing robust, interpretable, and clinically relevant AI-based assessment systems]]>
Copyrights © 2025
