Machine Learning for Violence Risk Assessment Using Dutch Clinical Notes

Research Articie March 04,2021

Pablo Mosteiro ¹ , Emil Rijcken ² , Kalliopi Zervanou ³ , Uzay Kaymak ⁴ , Floortje Scheepers ⁵ , Marco Spruit ⁶ hide author's information

1 : Utrecht University, Utrecht, The Netherlands
2 : Eindhoven University of Technology, Eindhoven, The Netherlands
3 : Eindhoven University of Technology, Eindhoven, The Netherlands
4 : Eindhoven University of Technology, Eindhoven, The Netherlands
5 : University Medical Center Utrecht, Utrecht, The Netherlands
6 : Utrecht University, Utrecht, The Netherlands; Leiden University Medical Center, Leiden, The Netherlands; Leiden Institute of Advanced Computer Science, Leiden, The Netherlands

DOI: https://doi.org/10.2991/jaims.d.210225.001

Keywords: Natural language processing; Topic modeling; Electronic health records; BERT; Evaluation metrics; Interpretability; Document classification; LDA; Random forests

Cite this article: Mosteiro P, Rijcken E, Zervanou K, Kaymak U, Scheepers F, Spruit M. Machine Learning for Violence Risk Assessment Using Dutch Clinical Notes. JAIMS [Internet]. 2021;2(1-2):44-5.

Full Text

PDF

Download Citation

Abstract

Violence risk assessment in psychiatric institutions enables interventions to avoid violence incidents. Clinical notes written by practitioners and available in electronic health records are valuable resources capturing unique information, but are seldom used to their full potential. We explore conventional and deep machine learning methods to assess violence risk in psychiatric patients using practitioner notes. The performance of our best models is comparable to the currently used questionnaire-based method, with an area under the Receiver Operating Characteristic curve of approximately 0.8. We find that the deep-learning model BERTje performs worse than conventional machine learning methods. We also evaluate our data and our classifiers to understand the performance of our models better. This is particularly important for the applicability of evaluated classifiers to new data, and is also of great interest to practitioners, due to the increased availability of new data in electronic format.