A neuro-symbolic method for understanding free-text medical evidence.
J Am Med Inform Assoc
; 28(8): 1703-1711, 2021 07 30.
Article
in English
| MEDLINE | ID: covidwho-1217859
ABSTRACT
OBJECTIVE:
We introduce Medical evidence Dependency (MD)-informed attention, a novel neuro-symbolic model for understanding free-text clinical trial publications with generalizability and interpretability. MATERIALS ANDMETHODS:
We trained one head in the multi-head self-attention model to attend to the Medical evidence Ddependency (MD) and to pass linguistic and domain knowledge on to later layers (MD informed). This MD-informed attention model was integrated into BioBERT and tested on 2 public machine reading comprehension benchmarks for clinical trial publications Evidence Inference 2.0 and PubMedQA. We also curated a small set of recently published articles reporting randomized controlled trials on COVID-19 (coronavirus disease 2019) following the Evidence Inference 2.0 guidelines to evaluate the model's robustness to unseen data.RESULTS:
The integration of MD-informed attention head improves BioBERT substantially in both benchmark tasks-as large as an increase of +30% in the F1 score-and achieves the new state-of-the-art performance on the Evidence Inference 2.0. It achieves 84% and 82% in overall accuracy and F1 score, respectively, on the unseen COVID-19 data.CONCLUSIONS:
MD-informed attention empowers neural reading comprehension models with interpretability and generalizability via reusable domain knowledge. Its compositionality can benefit any transformer-based architecture for machine reading comprehension of free-text medical evidence.Keywords
Full text:
Available
Collection:
International databases
Database:
MEDLINE
Main subject:
Natural Language Processing
/
Artificial Intelligence
/
Clinical Trials as Topic
/
Information Storage and Retrieval
/
Models, Neurological
Type of study:
Experimental Studies
/
Prognostic study
/
Randomized controlled trials
/
Reviews
Limits:
Humans
Language:
English
Journal:
J Am Med Inform Assoc
Journal subject:
Medical Informatics
Year:
2021
Document Type:
Article
Affiliation country:
Jamia
Similar
MEDLINE
...
LILACS
LIS