PSYCOP

The PSYCOP project aims at using the wealth of information in Electronic Health Records to improve patient care and treatment of those with mental illness.

To do this, we develop machine learning models with a specific focus on evaluation and clinical applicability.

We are currently developing tools for thoroughly validation our data and models, and quickly iterating on classical machine learning models. Check out our Github repositories for feature generation and data validation and model training and evaluation.

Ongoing and planned projects include:

• Representation learning on the multi-modal EHR data
• Deep sequential models for EHR time-series

For more information, see our paper which outlines the cohort and research directions, or reach out!

Members

Faculty:

• Professor Søren Dinesen Østergaard (PI)
• Andreas Aalkjær Danielsen, MD
• Associate professor Kristoffer Laigaard Nielbo

PhD students:

• Lasse Hansen
• Kenneth Enevoldsen
• Martin Bernstorff
• Erik Perfalk
• Frida Hæstrup