Exploring Domain-Sensitive Features for Extractive Summarization in the Medical Domain (original) (raw)
Related papers
Extractive summarization of medical documents using domain knowledge and corpus statistics
Australasian Medical Journal, 2012
Evidence Based Medicine (EBM) practice requires practitioners to extract evidence from published medical research when answering clinical queries. Due to the time-consuming nature of this practice, there is a strong motivation for systems that can automatically summarise medical documents and help practitioners find relevant information.
Enhancing Performance in Medical Articles Summarization with Multi-Feature Selection
International Journal of Electrical and Computer Engineering (IJECE), 2018
The research aimed at providing an outcome summary of extraordinary events information for public health surveillance systems based on the extraction of online medical articles. The data set used is 7,346 pieces. Characteristics possessed by online medical articles include paragraphs that comprise more than one and the core location of the story or important sentences scattered at the beginning, middle and end of a paragraph. Therefore, this study conducted a summary by maintaining important phrases related to the information of extraordinary events scattered in every paragraph in the medical article online. The summary method used is maximal marginal relevance with an n-best value of 0.7. While the multi feature selection in question is the use of features to improve the performance of the summary system. The first feature selection is the use of title and statistic number of word and noun occurrence, and weighting tf-idf. In addition, other features are word level category in medical content patterns to identify important sentences of each paragraph in the online medical article. The important sentences defined in this study are classified into three categories: core sentence, explanatory sentence, and supporting sentence. The system test in this study was divided into two categories, such as extrinsic and intrinsic test. Extrinsic test is comparing the summary results of the decisions made by the experts with the output resulting from the system. While intrinsic test compared three n-Best weighting value method, feature selection combination, and combined feature selection combination with word level category in medical content. The extrinsic evaluation result was 72%. While intrinsic evaluation result of feature selection combination merger method with word category in medical content was 91,6% for precision, 92,6% for recall and f-measure was 92,2%. Keyword: Text summarization Feature selection N-Best Second opinion Weighting Word level category in medical content
Modeling Medical Content for Automated Summarization
Annals of the New York Academy of Sciences, 2002
A BSTRACT : Medical information is available from a variety of new online resources. Given the number and diversity of sources, methods must be found that will enable users to quickly assimilate and determine the content of a document. Summarization is one such tool that can help users to quickly determine the main points of a document. Previous methods to automatically summarize text documents typically do not attempt to infer or define the content of a document. Rather these systems rely on secondary features or clues that may point to content. This paper describes text summarization techniques that enable users to focus on the key content of a document. The techniques presented here analyze groups of similar documents in order to form a content model. The content model is used to select sentences forming the summary. The technique does not require additional knowledge sources; thus the method should be applicable to any set of text documents.
Automated Medical Text Summarisation to Support Evidence-based Medicine
Clinical guidelines urge medical practitioners to perform Evidence-based Medicine: a practice that requires practitioners to incorporate the best available evidence from published research and from clinical practice, when making decisions. Due to the abundance of published medical research available, practitioners often fail to follow evidence-based guidelines at point-of-care due to time constraints. As such, this practice can vastly benefit from automatic systems that can generate short, reliable, evidence-based summaries in response to queries posed by the practitioners. Analysis of a corpus specialised for text summarisation in the evidence-based medicine domain suggests that a summarisation system tailored to this domain must be capable of performing two tasks: (i) query-focused text summarisation and (ii) automatic appraisal of the evidence. In this thesis, we utilise data from a specialised corpus to address these two facets of the problem. Our investigations lead to the following three contributions/findings: 1. A model for the automatic generation of evidence grades: We show that a supervised classification model can be used to perform automatic quality grading of evidence on a chosen scale. Rule-based approaches can be applied to the text of the medical articles to extract useful features, which can be utilised in the classification task. Using a sequence of high precision machine learning classifiers, we achieve recall and grading accuracies that are comparable to human performance, and significantly better than baseline systems.
A Novel System for Extractive Clinical Note Summarization using
Proceedings of the 2nd Clinical Natural Language Processing Workshop
While much data within a patient's electronic health record (EHR) is coded, crucial information concerning the patient's care and management remain buried in unstructured clinical notes, making it difficult and time-consuming for physicians to review during their usual clinical workflow. In this paper, we present our clinical note processing pipeline, which extends beyond basic medical natural language processing (NLP) with concept recognition and relation detection to also include components specific to EHR data, such as structured data associated with the encounter, sentencelevel clinical aspects, and structures of the clinical notes. We report on the use of this pipeline in a disease-specific extractive text summarization task on clinical notes, focusing primarily on progress notes by physicians and nurse practitioners. We show how the addition of EHR-specific components to the pipeline resulted in an improvement in our overall system performance and discuss the potential impact of EHR-specific components on other higher-level clinical NLP tasks.
Extractive evidence based medicine summarisation based on sentence-specific statistics
2012
Abstract We present an approach for extracting 3-sentence evidence-based summaries relevant to clinical questions. We approach this task as one of query-focused, extractive, single-document summarisation using sentence-specific statistics for each target sentence. We incorporate simple statistics and domain knowledge and show that such an approach is effective for identifying informative sentences from medical abstracts.
2005
Current text summarization strategies often draw upon one specific type of criteria to locate summary relevant text passages. For instance, they are statistical, discourse structure-based, or positional. In this paper, we argue that in order to arrive at an optimal summary, the whole range of linguistic criteria must be taken into account: textual, lexical, discursive, informative, and syntactic. First preliminary experiments carried out with medical articles in Spanish suggest the validity of our argumentation.
Text summarization in the biomedical domain: A systematic review of recent research
Journal of Biomedical Informatics, 2014
Objective: The amount of information for clinicians and clinical researchers is growing exponentially. Text summarization reduces information as an attempt to enable users to find and understand relevant source texts more quickly and effortlessly. In recent years, substantial research has been conducted to develop and evaluate various summarization techniques in the biomedical domain. The goal of this study was to systematically review recent published research on summarization of textual documents in the biomedical domain. Materials and methods: MEDLINE (2000 to October 2013), IEEE Digital Library, and the ACM digital library were searched. Investigators independently screened and abstracted studies that examined text summarization techniques in the biomedical domain. Information is derived from selected articles on five dimensions: input, purpose, output, method and evaluation. Results: Of 10,786 studies retrieved, 34 (0.3%) met the inclusion criteria. Natural language processing (17; 50%) and a hybrid technique comprising of statistical, Natural language processing and machine learning (15; 44%) were the most common summarization approaches. Most studies (28; 82%) conducted an intrinsic evaluation. Discussion: This is the first systematic review of text summarization in the biomedical domain. The study identified research gaps and provides recommendations for guiding future research on biomedical text summarization. Conclusion: Recent research has focused on a hybrid technique comprising statistical, language processing and machine learning techniques. Further research is needed on the application and evaluation of text summarization in real research or patient care settings.
arXiv (Cornell University), 2024
This work presents a dynamic vocabulary adaptation strategy, MEDVOC, for fine-tuning pretrained language models (PLMs) like BertSum-Abs, BART, and PEGASUS for improved medical text summarization. In contrast to existing domain adaptation approaches in summarization, MEDVOC treats vocabulary as an optimizable parameter and optimizes the PLM vocabulary based on fragment score conditioned only on the downstream task's reference summaries. Unlike previous works on vocabulary adaptation (limited only to classification tasks), optimizing vocabulary based on summarization tasks requires an extremely costly intermediate fine-tuning step on large summarization datasets. To that end, our novel fragment score-based hyperparameter search very significantly reduces this fine-tuning timefrom 450 days to less than 2 days on average. Furthermore, while previous works on vocabulary adaptation are often primarily tied to single PLMs, MEDVOC is designed to be deployable across multiple PLMs (with varying model vocabulary sizes, pre-training objectives, and model sizes)-bridging the limited vocabulary overlap between the biomedical literature domain and PLMs. MEDVOC outperforms baselines by 15.74% in terms of Rouge-L in zero-shot setting and shows gains of 17.29% in high Out-Of-Vocabulary (OOV) concentrations. Our human evaluation shows MEDVOC generates more faithful medical summaries (88% compared to 59% in baselines).