A systematic review of automatic text summarization for biomedical literature and EHRs

Abstract

Objective Biomedical text summarization helps biomedical information seekers avoid information overload by reducing the length of a document while preserving the contents’ essence. Our systematic review investigates the most recent biomedical text summarization researches on biomedical literature and electronic health records by analyzing their techniques, areas of application, and evaluation methods. We identify gaps and propose potential directions for future research. Materials and Methods This review followed the PRISMA methodology and replicated the approaches adopted by the previous systematic review published on the same topic. We searched 4 databases (PubMed, ACM Digital Library, Scopus, and Web of Science) from January 1, 2013 to April 8, 2021. Two reviewers independently screened title, abstract, and full-text for all retrieved articles. The conflicts were resolved by the third reviewer. The data extraction of the included articles was in 5 dimensions: input, purpose, output, method, and evaluation. Results Fifty-eight out of 7235 retrieved articles met the inclusion criteria. Thirty-nine systems used single-document biomedical research literature as their input, 17 systems were explicitly designed for clinical support, 47 systems generated extractive summaries, and 53 systems adopted hybrid methods combining computational linguistics, machine learning, and statistical approaches. As for the assessment, 51 studies conducted an intrinsic evaluation using predefined metrics. Discussion and Conclusion This study found that current biomedical text summarization systems have achieved good performance using hybrid methods. Studies on electronic health records summarization have been increasing compared to a previous survey. However, the majority of the works still focus on summarizing literature.

Monogenetic rare diseases in biomedical databases and text mining

“…Automatic text mining reduces time and effort and gets maximum results from published medical texts, case-reports, experimental data, patents, and systematic reviews. There are many technologies for automatic text reading that may achieve high quality of extracting facts or citations from articles in a simplified and structured way (Mishra et al, 2014; Wang et al, 2021).…”

Increasing Women’s Knowledge about HPV Using BERT Text Summarization: An Online Randomized Study

“…The evaluation of text summarization techniques can be intrinsic or extrinsic. Using intrinsic evaluation, the quality of the summarized text is assessed based on metrics or attributes, such as ROUGE, accuracy, comprehensiveness, and relevancy [9,36]. The extrinsic evaluation assesses the summaries by applying them to user-related tasks, such as information seeking, to measure the user's efficiency, comprehension, success rate, or decision-making accuracy [9,36].…”

“…Using intrinsic evaluation, the quality of the summarized text is assessed based on metrics or attributes, such as ROUGE, accuracy, comprehensiveness, and relevancy [9,36]. The extrinsic evaluation assesses the summaries by applying them to user-related tasks, such as information seeking, to measure the user's efficiency, comprehension, success rate, or decision-making accuracy [9,36]. It is concerned with evaluating how effective the automatically generated summaries are in helping users perform their normal tasks in realistic contexts to achieve their desired goals.…”

“…It is concerned with evaluating how effective the automatically generated summaries are in helping users perform their normal tasks in realistic contexts to achieve their desired goals. Most of the evaluations of existing summarization techniques, including BERT, have used intrinsic evaluations, which focus on the components of the text summarization method and ignore the impact of text summarization in real-world settings [9,36,37]. However, for better maturity of the text summarization field in the medical and health informatics domain, more studies involving users need to be conducted.…”

“…One fundamental challenge to women's ability to acquire the necessary knowledge and awareness of HPV is the large amount of information. Machine learning techniques, such as deep learning to reduce information overload and focusing on the most valuable information through text summarization, have been validated in previous studies [9]. To the best of our knowledge, deep learning has not yet been used in the context of increasing awareness of, and knowledge about, HPV.…”