Assertion Detection in Clinical Notes: Medical Language Models to the Rescue?

Betty van Aken, Ivana Trajanovska, Amy Siu, Manuel Mayrdorfer, Klemens Budde, Alexander Loeser


Abstract
In order to provide high-quality care, health professionals must efficiently identify the presence, possibility, or absence of symptoms, treatments and other relevant entities in free-text clinical notes. Such is the task of assertion detection - to identify the assertion class (present, possible, absent) of an entity based on textual cues in unstructured text. We evaluate state-of-the-art medical language models on the task and show that they outperform the baselines in all three classes. As transferability is especially important in the medical domain we further study how the best performing model behaves on unseen data from two other medical datasets. For this purpose we introduce a newly annotated set of 5,000 assertions for the publicly available MIMIC-III dataset. We conclude with an error analysis that reveals situations in which the models still go wrong and points towards future research directions.
Anthology ID:
2021.nlpmc-1.5
Volume:
Proceedings of the Second Workshop on Natural Language Processing for Medical Conversations
Month:
June
Year:
2021
Address:
Online
Venues:
NAACL | NLPMC
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
35–40
Language:
URL:
https://aclanthology.org/2021.nlpmc-1.5
DOI:
10.18653/v1/2021.nlpmc-1.5
Bibkey:
Cite (ACL):
Betty van Aken, Ivana Trajanovska, Amy Siu, Manuel Mayrdorfer, Klemens Budde, and Alexander Loeser. 2021. Assertion Detection in Clinical Notes: Medical Language Models to the Rescue?. In Proceedings of the Second Workshop on Natural Language Processing for Medical Conversations, pages 35–40, Online. Association for Computational Linguistics.
Cite (Informal):
Assertion Detection in Clinical Notes: Medical Language Models to the Rescue? (van Aken et al., NLPMC 2021)
Copy Citation:
PDF:
https://aclanthology.org/2021.nlpmc-1.5.pdf
Code
 bvanaken/clinical-assertion-data