Wrist Angel - hug mountain


We aim to improve assessment and psychotherapy for pediatric obsessive-compulsive disorder (OCD) through developing artificial intelligence tools to support patients, parents and therapists in cognitive behavioural therapy.


OCD is a chronic and debilitating psychiatric disorder, which can damage self-esteem, create conflict within families, make it difficult to concentrate, attend school and maintain friends and even shorten life-expectancy [1,2]. Yet, prevention and early intervention efforts have been neglected in OCD [3]. Clinical guidelines recommend either individual or family-based cognitive behavioral therapy (CBT) as the first-line treatment for OCD [4]. However, children and adolescents often do not complete CBT homework assignments due to lack of understanding, engagement or remembering. Additionally, parents often struggle to support their children and inadvertently accommodate their children’s symptoms by enabling avoidance during exposures.


OCD is like a devil on your shoulder filling your head with disturbing thoughts, images, worry and doubt, telling you to perform acts to avoid discomfort or bad things from happening. The OCD devil follows patients around everywhere and screams the loudest when there are many external stressors in the patient’s life. Resisting OCD can be especially difficult for youth outside of therapy sessions and would be easier if they had an angel on their wrist, the Wrist Angel, to counteract the OCD devil. Wrist Angel is a wearable sensor and mobile app that will predict an OCD-episode and ultimately provide youth with OCD with personalized therapy interventions before distress levels escalate to levels at which therapeutic learning is less likely to occur.

Given the tech-savviness of much of today’s youth, a digital tool may help children engage more actively in treatment. Thus, we will use AI tools to improve the quality, flexibility, and efficiency of psychotherapy. AI tools can support learning, reduce manual labour and nudge skills practice, thereby increasing compliance. A wearable sensor would automate distress monitoring and passively collect objective measures, which is the first step in personalizing treatment. The AI will use passively collected physiological data and actively collected clinical and biochemical data from patients and parents to detect and predict episodes of OCD-related stress. Subsequently, the AI tools will be developed to monitor patient progress, identify the optimal timing and need for specific interventions and aid treatment planning. 

This study is affiliated with the TECTO-study.

For information in Danish, please see our the Wrist Angel webpage on the Child and Adolescent Mental Health Centre’s website.


  1. Gillan C, Fineberg N, Robbins T. A trans-diagnostic perspective on obsessive-compulsive disorder. Psychological Medicine. 2017;47(9):1528-1548.

  2. Meier S, Mattheisen M, Mors O, Schendel D, Mortensen P, Plessen K. Mortality Among Persons With Obsessive-Compulsive Disorder in Denmark. JAMA Psychiatry. 2016;73(3):268-274.

  3. Stein D, Costa D, Lochner C, Miguel E, Reddy Y, Shavitt R et al. Obsessive–compulsive disorder. Nature Reviews Disease Primers. 2019;5(1):52.

  4. Sundhedsstyrelsen. NKR: Behandling af obsessiv-kompulsiv tilstand (OCD) [Internet]. Sst.dk. 2021. Available from: https://www.sst.dk/da/udgivelser/2016/nkr-behandling-af-ocd


Sneha Das
DTU Compute
+45 91 44 08 64


Jakob Eyvind Bardram
Head of Sections, Professor
DTU Health Tech
+45 45 25 53 11


Line Katrine Harder Clemmensen
Associate Professor
DTU Compute
+45 45 25 37 64


Nicole Nadine Lønfeldt
Senior Scientist
Child and Adolescent Mental Health Centre, Region Hovedstaden
+45 38 64 11 48


Anne Katrine Pagsberg
Head of Department, Professor
Child and Adolescent Mental Health Centre, Region Hovedstaden
+45 38 64 12 48
24 MAY 2022