Enhancing PIO Element Detection in Medical Text Using Contextualized Embedding
This work addresses the problem of extracting Population, Intervention, and Outcome elements from medical texts for Evidence Based Medicine, representing an incremental improvement through dataset refinement and embedding optimization.
The paper tackles PIO element detection in medical text by building an improved training dataset with reduced redundancy and ambiguity, and by developing a multi-label classifier using BERT embeddings. The results show that domain-specific pre-trained embeddings optimize performance, with ensemble methods and boosting techniques providing further enhancements when features are adequately chosen.
In this paper, we investigate a new approach to Population, Intervention and Outcome (PIO) element detection, a common task in Evidence Based Medicine (EBM). The purpose of this study is two-fold: to build a training dataset for PIO element detection with minimum redundancy and ambiguity and to investigate possible options in utilizing state of the art embedding methods for the task of PIO element detection. For the former purpose, we build a new and improved dataset by investigating the shortcomings of previously released datasets. For the latter purpose, we leverage the state of the art text embedding, Bidirectional Encoder Representations from Transformers (BERT), and build a multi-label classifier. We show that choosing a domain specific pre-trained embedding further optimizes the performance of the classifier. Furthermore, we show that the model could be enhanced by using ensemble methods and boosting techniques provided that features are adequately chosen.