Digital behavioural tests as diagnostic aid for psychosis

Timely interventions have a proven benefit for people experiencing psychotic illness. One bottleneck to accessing timely interventions is the referral process to the specialist team for early psychosis (STEP). Many general practitioners lack awareness or confidence in recognising psychotic symptoms or state. Additionally, referrals for people without apparent psychotic symptoms, although beneficial at a population level, lead to excessive workload for STEPs. There is a clear unmet need for accurate stratification of STEPs users and healthy cohorts. Here we propose a new approach to addressing this need via the application of digital behavioural tests. To discriminate between the STEPs users (SU; n=32) and controls (n=32, age and sex matched), we employed k-nearest neighbours (kNN) classifier, and applied it to objective, quantitative and interpretable features derived from the mirror game (MG) and trail making task (TMT). The MG is a movement coordination task shown to be a potential socio-motor biomarker of schizophrenia, while TMT is a neuropsychiatric test of cognitive function. We show that the proposed classifier achieves an excellent performance, AUC = 0.89 (95%CI 0.73-1), Sensitivity = 0.75 (95%CI 0.5-1), Specificity = 1 (95%CI 0.62-1), evaluated on 25% hold-out and 1000 folds. We demonstrate that this performance is underpinned by the large effect sizes of the differences between the cohorts in terms of the features used for classification. We also find that MG and TMT are unsuitable in isolation to successfully differentiate between SU with and without at-risk-mental-state or first episode psychosis with sufficient level of performance. Our findings show that introduction of standardised battery of digital behavioural tests could benefit both clinical and research practice. Including digital behavioural tests into healthcare practice could allow precise phenotyping and stratification of the highly heterogenous population of people referred to STEPs resulting in quicker and more personalised diagnosis. Moreover, the high specificity of digital behavioural tests could facilitate the identification of more homogeneous clinical high-risk populations, benefiting research on prognostic instruments for psychosis. In summary, our study demonstrates that cheap off-the-shelf equipment (laptop computer and a leap motion sensor) can be used to record clinically relevant behavioural data that could be utilised in digital mental health applications. Author summaryNeuropsychiatric assessment and accurate diagnosis are notoriously challenging. Psychosis represents a classical example of this challenge where many at-risk of psychotic illness individuals (often very young) are misdiagnosed and/or inappropriately treated clinically. Our study demonstrates that combining digital tests with data analytics has potential for simplifying neuropsychiatric assessment. It shows that using measurements from TMT and MG allows to differentiate between people accepted for assessment in specialist team for early psychosis (STEP) and controls with excellent performance (AUROC > 0.9), while achieving 100% specificity (no false positive detections). The study shows feasibility of using cheap, portable equipment, assembled from off-the-shelf components, for collection of clinically relevant data that could be used to inform clinical decision making. Moreover, our study, with its state-of-the-art performance and interpretable results, demonstrate high clinical potential of implementing digital batteries of behavioural tests in clinical practice. Such developments would not only help to stratify STEPs users but would facilitate rapid assessment for all people seeking care in early intervention services. This in turn would contribute to improving the quality of life and wellbeing of individuals at risk of developing psychosis. FundingEPSRC Impact Acceleration Account, Impact & Knowledge Exchange Award, Jean Golding Institute seed corn, Avon & Wiltshire Mental Health Partnership NHS Trust Research Capability Funding. PS was generously supported by the Wellcome Trust Institutional Strategic Support Award 204909/Z/16/Z. KTA gratefully acknowledges the financial support of the EPSRC via grant EP/T017856/1. For the purpose of open access, the authors have applied a Creative Commons Attribution (CC BY) licence to any Author Accepted Manuscript version arising.