Hybrid methods for improving information access in clinical documents: concept, assertion, and relation identification

Minard, Anne-Lyse; Ligozat, Anne-Laure; Abacha, Asma Ben; Bernhard, Delphine; Cartoni, Bruno; Deléger, Louise; Grau, Brigitte; Rosset, Sophie; Zweigenbaum, Pierre; Grouin, Cyril

doi:10.1136/amiajnl-2011-000154

Cited by 35 publications

(26 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Most criteria were words or UMLS CUIs that describe facial fractures or orbital anatomy. Support Vector Machines (SVM) have been previously shown to be very successful in text classification, including the medical domain, but provide no way of verifying the sensibility of the algorithm. Although we planned to perform only decision trees as a transparent form of machine learning, we did confirm that the results of classification using SVM (using the WEKA SMO classifier) showed similar performance…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Automated Outcome Classification of Emergency Department Computed Tomography Imaging Reports

et al. 2013

View full text Add to dashboard Cite

Background Reliably abstracting outcomes from free-text electronic medical records remains a challenge. While automated classification of free text has been a popular medical informatics topic, performance validation using real-world clinical data has been limited. The two main approaches are linguistic (natural language processing [NLP]) and statistical (machine learning). The authors have developed a hybrid system for abstracting computed tomography (CT) reports for specified outcomes. Objectives The objective was to measure performance of a hybrid NLP and machine learning system for automated outcome classification of emergency department (ED) CT imaging reports. The hypothesis was that such a system is comparable to medical personnel doing the data abstraction. Methods A secondary analysis was performed on a prior diagnostic imaging study on 3,710 blunt facial trauma victims. Staff radiologists dictated CT reports as free text, which were then deidentified. A trained data abstractor manually coded the reference standard outcome of acute orbital fracture, with a random subset double-coded for reliability. The data set was randomly split evenly into training and testing sets. Training patient reports were used as input to the Medical Language Extraction and Encoding (MedLEE) NLP tool to create structured output containing standardized medical terms and modifiers for certainty and temporal status. Findings were filtered for low certainty and past/future modifiers and then combined with the manual reference standard to generate decision tree classifiers using data mining tools Waikato Environment for Knowledge Analysis (WEKA) 3.7.5 and Salford Predictive Miner 6.6. Performance of decision tree classifiers was evaluated on the testing set with or without NLP processing. Results The performance of machine learning alone was comparable to prior NLP studies (sensitivity = 0.92, specificity = 0.93, precision = 0.95, recall = 0.93, f-score = 0.94), and the combined use of NLP and machine learning shows further improvement (sensitivity = 0.93, specificity = 0.97, precision = 0.97, recall = 0.96, f-score = 0.97). This performance is similar to, or better than, that of medical personnel in previous studies. Conclusions A hybrid NLP and machine learning automated classification system shows promise in coding free-text electronic clinical data.

show abstract

Section: Discussionmentioning

confidence: 99%

“…Regardless, more generalizable automated outcome classification pairing NLP software and machine learning techniques are now possible. This approach has shown to have the potential to code EHR data, although most prior studies have been performed on documents mocked up for NLP testing and never validated on real‐world data.…”

mentioning

confidence: 99%

Automated Outcome Classification of Emergency Department Computed Tomography Imaging Reports

et al. 2013

View full text Add to dashboard Cite

show abstract

“…After preliminary experiment we found that we did not have For extracting relations among entities we considered all sentences having more than one entities in each discharge summary to check whether any relation exists between them or not. In our experiment we assume that entities and their types are already known like other existing works (Rink et al, 2011;Minard et al, 2011a;Minard et al, 2011b). We created data sample for every pair of entities present in the sentence and labeled it with the existing relation type.…”

Section: Dataset and Experimental Settingsmentioning

confidence: 99%

Relation extraction from clinical texts using domain invariant convolutional neural network

Sahu¹,

Anand²,

Oruganty³

et al. 2016

Proceedings of the 15th Workshop on Biomedical Natural Language Processing

View full text Add to dashboard Cite

In recent years extracting relevant information from biomedical and clinical texts such as research articles, discharge summaries, or electronic health records have been a subject of many research efforts and shared challenges. Relation extraction is the process of detecting and classifying the semantic relation among entities in a given piece of texts. Existing models for this task in biomedical domain use either manually engineered features or kernel methods to create feature vector. These features are then fed to classifier for the prediction of the correct class. It turns out that the results of these methods are highly dependent on quality of user designed features and also suffer from curse of dimensionality. In this work we focus on extracting relations from clinical discharge summaries. Our main objective is to exploit the power of convolution neural network (CNN) to learn features automatically and thus reduce the dependency on manual feature engineering. We evaluate performance of the proposed model on i2b2-2010 clinical relation extraction challenge dataset. Our results indicate that convolution neural network can be a good model for relation exaction in clinical text without being dependent on expert's knowledge on defining quality features.

show abstract

“…39 Classification is the most common method in textual analysis; this can be done using Perceptron, Support Vector Machine (SVM), Naive Bayes (NB), k-NN, Maximum Entropy Model (MEM), and Decision Tree (DT). 42 Optimal features need to be defined for the selected classifier, including sentencelevel, entity-level, and token-level features, as shown in Table 1.…”

Section: Semantic Recognition Layermentioning

confidence: 99%

Framework for Automatic Semantic Annotation of Arabic Websites

Helmy

Albukhitan

2016

Int. J. Coop. Info. Syst.

View full text Add to dashboard Cite

In order to achieve the vision of the semantic Web, it is important to have enough amount of semantic content on the Web sources. To produce the semantic content on the existing Web, semantic annotation of the Web sources is required. Semantic annotation adds machine-readable content to the Web sources. Because the Web is growing at an exponential rate, semantic annotation by hand is not possible. In this paper, we present an Automatic Semantic Annotation Framework (ASAF) for semantic annotation of Arabic Web sources based on the domain ontologies. We present a learning approach that utilizes public Arabic resources, such as Wikipedia and WordNet for building Arabic ontologies. Moreover, we present different approaches for extracting name entities and relationships from Arabic Web sources. As a case study, we have developed and expanded a set of Arabic ontologies related to food, health, and nutrition through a set of processes. We have also developed the ASAF prototype, and showed how it can utilize these ontologies for extracting health, food related name entities, and relationships from the Web sources in order to annotate and store them in the knowledge-base. We conducted several experiments to test the capability of ASAF in recognizing the name entities and relationships using different approaches. Empirical evaluations of ASAF show promising performance results in terms of precision, recall, and F -measure. The outcome of the presented framework could be utilized by semantic Web searching applications to retrieve precise answers to the end user smarter queries. An important feature of ASAF is that it could be ported to other domains with minimal extension. ASAF also contributes to the vision of the semantic Web in the target domains in Arabic Web sources.

show abstract

Hybrid methods for improving information access in clinical documents: concept, assertion, and relation identification

Cited by 35 publications

References 17 publications

Automated Outcome Classification of Emergency Department Computed Tomography Imaging Reports

Automated Outcome Classification of Emergency Department Computed Tomography Imaging Reports

Relation extraction from clinical texts using domain invariant convolutional neural network

Framework for Automatic Semantic Annotation of Arabic Websites

Contact Info

Product

Resources

About