Dina Demner-Fushman

Methodological Review: What can natural language processing do for clinical decision support?

Computerized clinical decision support (CDS) aims to aid decision making of health care providers and the public by providing easily accessible health-related information at the point and time it is needed. natural language processing (NLP) is instrumental in using free-text information to drive CDS, representing clinical knowledge and CDS interventions in standardized formats, and leveraging clinical narrative. The early innovative NLP research of clinical narrative was followed by a period of stable research conducted at the major clinical centers and a shift of mainstream interest to biomedical NLP. This review primarily focuses on the recently renewed interest in development of fundamental NLP methods and advances in the NLP systems for CDS. The current solutions to challenges posed by distinct sublanguages, intended user groups, and support goals are discussed.

Essie: A Concept-based Search Engine for Structured Biomedical Text

This article describes the algorithms implemented in the Essie search engine that is currently serving several Web sites at the National Library of Medicine. Essie is a phrase-based search engine with term and concept query expansion and probabilistic relevancy ranking. Essies design is motivated by an observation that query terms are often conceptually related to terms in a document, without actually occurring in the document text. Essies performance was evaluated using data and standard evaluation methods from the 2003 and 2006 Text REtrieval Conference (TREC) Genomics track. Essie was the best-performing search engine in the 2003 TREC Genomics track and achieved results comparable to those of the highest-ranking systems on the 2006 TREC Genomics track task. Essie shows that a judicious combination of exploiting document structure, phrase searching, and concept based query expansion is a useful approach for information retrieval in the biomedical domain.

Extracting semantic predications from Medline citations for pharmacogenomics.

We describe a natural language processing system (Enhanced SemRep) to identify core assertions on pharmacogenomics in Medline citations. Extracted information is represented as semantic predications covering a range of relations relevant to this domain. The specific relations addressed by the system provide greater precision than that achievable with methods that rely on entity co-occurrence. The development of Enhanced SemRep is based on the adaptation of an existing system and crucially depends on domain knowledge in the Unified Medical Language System. We provide a preliminary evaluation (55% recall and 73% precision) and discuss the potential of this system in assisting both clinical practice and scientific investigation.

Automatic summarization of MEDLINE citations for evidence-based medical treatment: A topic-oriented evaluation

As the number of electronic biomedical textual resources increases, it becomes harder for physicians to find useful answers at the point of care. Information retrieval applications provide access to databases; however, little research has been done on using automatic summarization to help navigate the documents returned by these systems. After presenting a semantic abstraction automatic summarization system for MEDLINE citations, we concentrate on evaluating its ability to identify useful drug interventions for 53 diseases. The evaluation methodology uses existing sources of evidence-based medicine as surrogates for a physician-annotated reference standard. Mean average precision (MAP) and a clinical usefulness score developed for this study were computed as performance metrics. The automatic summarization system significantly outperformed the baseline in both metrics. The MAP gain was 0.17 (p<0.01) and the increase in the overall score of clinical usefulness was 0.39 (p<0.05).

Towards Automatic Recognition of Scientifically Rigorous Clinical Research Evidence

The growing numbers of topically relevant biomedical publications readily available due to advances in document retrieval methods pose a challenge to clinicians practicing evidence-based medicine. It is increasingly time consuming to acquire and critically appraise the available evidence. This problem could be addressed in part if methods were available to automatically recognize rigorous studies immediately applicable in a specific clinical situation. We approach the problem of recognizing studies containing useable clinical advice from retrieved topically relevant articles as a binary classification problem. The gold standard used in the development of PubMed clinical query filters forms the basis of our approach. We identify scientifically rigorous studies using supervised machine learning techniques (Naïve Bayes, support vector machine (SVM), and boosting) trained on high-level semantic features. We combine these methods using an ensemble learning method (stacking). The performance of learning methods is evaluated using precision, recall and F(1) score, in addition to area under the receiver operating characteristic (ROC) curve (AUC). Using a training set of 10,000 manually annotated MEDLINE citations, and a test set of an additional 2,000 citations, we achieve 73.7% precision and 61.5% recall in identifying rigorous, clinically relevant studies, with stacking over five feature-classifier combinations and 82.5% precision and 84.3% recall in recognizing rigorous studies with treatment focus using stacking over word + metadata feature vector. Our results demonstrate that a high quality gold standard and advanced classification methods can help clinicians acquire best evidence from the medical literature.

Lower short- and long-term mortality associated with overweight and obesity in a large cohort study of adult intensive care unit patients

IntroductionTwo thirds of United States adults are overweight or obese, which puts them at higher risk of developing chronic diseases and of death compared with normal-weight individuals. However, recent studies have found that overweight and obesity by themselves may be protective in some contexts, such as hospitalization in an intensive care unit (ICU). Our objective was to determine the relation between body mass index (BMI) and mortality at 30 days and 1 year after ICU admission.MethodsWe performed a cohort analysis of 16,812 adult patients from MIMIC-II, a large database of ICU patients at a tertiary care hospital in Boston, Massachusetts. The data were originally collected during the course of clinical care, and we subsequently extracted our dataset independent of the study outcome.ResultsCompared with normal-weight patients, obese patients had 26% and 43% lower mortality risk at 30 days and 1 year after ICU admission, respectively (odds ratio (OR), 0.74; 95% confidence interval (CI), 0.64 to 0.86) and 0.57 (95% CI, 0.49 to 0.67)); overweight patients had nearly 20% and 30% lower mortality risk (OR, 0.81; 95% CI, 0.70 to 0.93) and OR, 0.68 (95% CI, 0.59 to 0.79)). Severely obese patients (BMI ≥ 40 kg/m2) did not have a significant survival advantage at 30 days (OR, 0.94; 95% CI, 0.74 to 1.20), but did have 30% lower mortality risk at 1 year (OR, 0.70 (95% CI, 0.54 to 0.90)). No significant difference in admission acuity or ICU and hospital length of stay was found across BMI categories.ConclusionOur study supports the hypothesis that patients who are overweight or obese have improved survival both 30 days and 1 year after ICU admission.

Biomedical Text Mining: A Survey of Recent Progress

The biomedical community makes extensive use of text mining technology. In the past several years, enormous progress has been made in developing tools and methods, and the community has been witness to some exciting developments. Although the state of the community is regularly reviewed, the sheer volume of work related to biomedical text mining and the rapid pace in which progress continues to be made make this a worthwhile, if not necessary, endeavor. This chapter provides a brief overview of the current state of text mining in the biomedical domain. Emphasis is placed on the resources and tools available to biomedical researchers and practitioners, as well as the major text mining tasks of interest to the community. These tasks include the recognition of explicit facts from biomedical literature, the discovery of previously unknown or implicit facts, document summarization, and question answering. For each topic, its basic challenges and methods are outlined and recent and influential work is reviewed.

Design and Development of a Multimodal Biomedical Information Retrieval System

The search for relevant and actionable information is a key to achieving clinical and research goals in biomedicine. Biomedical information exists in different forms: as text and illustrations in journal articles and other documents, in images stored in databases, and as patients’ cases in electronic health records. This paper presents ways to move beyond conventional text-based searching of these resources, by combining text and visual features in search queries and document representation. A combination of techniques and tools from the fields of natural language processing, information retrieval, and content-based image retrieval allows the development of building blocks for advanced information services. Such services enable searching by textual as well as visual queries, and retrieving documents enriched by relevant images, charts, and other illustrations from the journal literature, patient records and image databases.

Automatically Evaluating Answers to Definition Questions

Following recent developments in the automatic evaluation of machine translation and document summarization, we present a similar approach, implemented in a measure called Pourpre, for automatically evaluating answers to definition questions. Until now, the only way to assess the correctness of answers to such questions involves manual determination of whether an information nugget appears in a systems response. The lack of automatic methods for scoring system output is an impediment to progress in the field, which we address with this work. Experiments with the TREC 2003 and TREC 2004 QA tracks indicate that rankings produced by our metric correlate highly with official rankings, and that Pourpre outperforms direct application of existing metrics.

Evaluating performance of biomedical image retrieval systems--an overview of the medical image retrieval task at ImageCLEF 2004-2013.

Medical image retrieval and classification have been extremely active research topics over the past 15 years. Within the ImageCLEF benchmark in medical image retrieval and classification, a standard test bed was created that allows researchers to compare their approaches and ideas on increasingly large and varied data sets including generated ground truth. This article describes the lessons learned in ten evaluation campaigns. A detailed analysis of the data also highlights the value of the resources created.