PSYCOP: The PSYchiatric Clinical Outcome Prediction cohort 

Mental illness poses significant challenges to both the individual and society, with reduced quality of life and lifespan being common among those affected. Precision psychiatry seeks to address these challenges by providing personalized predictions for the course of illness and response to treatment. However, studies lack external validation and practical implementation, limiting their impact in clinical practice.  

To overcome these limitations, the PSYCOP cohort was established as a retrospective study, including data from all patients with at least one contact with the psychiatric services of the Central Denmark Region between January 1, 2011, and October 28, 2020. The primary goal of PSYCOP is to develop robust prediction models for a diverse range of clinical outcomes, thereby paving the way for a future where precision psychiatry seamlessly integrates into mental health care as a reliable component in treatment.  

Role of Center for Humanities Computing  

To leverage machine learning methodologies for predicting clinical outcomes in precision psychiatry, Center for Humanities Computing has brought forth its expertise in effectively handling large data sets (i.e., electronic health records), developing specialized models, and offering state-of-the-art hardware and software resources. 

Creating prediction models 

Center for Humanities Computing has been involved in creating prediction models for a wide array of clinical outcomes using advanced machine learning techniques. Making it possible to utilize clinical variables such as medications and diagnoses, as well as information from free-text clinical notes. These specialized machine learning models include transformer-based language models, which enhance the accuracy and depth of the predictions.