EBM+: Advancing Evidence-Based Medicine via two level automatic identification of Populations, Interventions, Outcomes in medical literature
作者：
Highlights：
• Neural end-to-end PICO Entity Recognizer that identifies Population, Intervention/Comparator and Outcome entities in medical publications.
• Novel Neural Network architecture for Entity Recognition involving a self-attention mechanism, a 2D Convolution feature extraction from character embeddings and a Highway residual connection.
• A PICO Statement classifier that identifies sentences containing all the PICO Entities and answering clinical questions.
• A high quality, manually annotated by medical practitioners dataset for PICO Statement classification.
摘要
•Neural end-to-end PICO Entity Recognizer that identifies Population, Intervention/Comparator and Outcome entities in medical publications.•Novel Neural Network architecture for Entity Recognition involving a self-attention mechanism, a 2D Convolution feature extraction from character embeddings and a Highway residual connection.•A PICO Statement classifier that identifies sentences containing all the PICO Entities and answering clinical questions.•A high quality, manually annotated by medical practitioners dataset for PICO Statement classification.
论文关键词：Evidence Based Medicine,PICO,Machine learning,Neural networks,Natural Language Processing
论文评审过程：Received 18 February 2020, Revised 11 August 2020, Accepted 12 August 2020, Available online 13 August 2020, Version of Record 22 August 2020.
论文官网地址：https://doi.org/10.1016/j.artmed.2020.101949
