scholarly journals Effects of Negation and Uncertainty Stratification on Text-Derived Patient Profile Similarity

2021 ◽  
Vol 3 ◽  
Author(s):  
Luke T. Slater ◽  
Andreas Karwath ◽  
Robert Hoehndorf ◽  
Georgios V. Gkoutos
Keyword(s):  
Differential Diagnosis ◽  
Text Mining ◽  
Information Extraction ◽  
Semantic Similarity ◽  
Outcome Prediction ◽  
Primary Diagnosis ◽  
Patient Profile ◽  
Profile Similarity ◽  
Standard Component ◽  
Mimic Iii

Semantic similarity is a useful approach for comparing patient phenotypes, and holds the potential of an effective method for exploiting text-derived phenotypes for differential diagnosis, text and document classification, and outcome prediction. While approaches for context disambiguation are commonly used in text mining applications, forming a standard component of information extraction pipelines, their effects on semantic similarity calculations have not been widely explored. In this work, we evaluate how inclusion and disclusion of negated and uncertain mentions of concepts from text-derived phenotypes affects similarity of patients, and the use of those profiles to predict diagnosis. We report on the effectiveness of these approaches and report a very small, yet significant, improvement in performance when classifying primary diagnosis over MIMIC-III patient visits.

scholarly journals Evaluating Semantic Similarity Methods for Comparison of Text-derived Phenotype Profiles

2021 ◽  
Author(s):  
Luke T Slater ◽  
Sophie Russell ◽  
Silver Makepeace ◽  
Alexander Carberry ◽  
Andreas Karwath ◽  
...  
Keyword(s):  
Differential Diagnosis ◽  
Semantic Similarity ◽  
Large Body ◽  
Similarity Measures ◽  
Primary Diagnosis ◽  
Experimental Conditions ◽  
Clinical Text ◽  
Automated Coding ◽  
Protein Interaction Prediction ◽  
Mimic Iii

Semantic similarity is a valuable tool for analysis in biomedicine. When applied to phenotype profiles derived from clinical text, they have the capacity to enable and enhance 'patient-like me' analyses, automated coding, differential diagnosis, and outcome prediction, by leveraging the wealth of background knowledge provided by biomedical ontologies. While a large body of work exists exploring the use of semantic similarity for multiple tasks, including protein interaction prediction, and rare disease differential diagnosis, there is less work exploring comparison of patient phenotype profiles for clinical tasks. Moreover, there are no experimental explorations of optimal parameters or methods in the area. In this work, we develop a reproducible platform for benchmarking experimental conditions for patient phentoype similarity. Using the platform, we evaluate the task of ranking shared primary diagnosis from uncurated phenotype profiles derived from text narrative associated with admissions in MIMIC-III. In doing this, we identify and interpret the performance of a large number of semantic similarity measures for this task, and provide a basis for further research on related tasks in the area.

Mining information from sentences through Semantic Web data and Information Extraction tasks

2020 ◽  
pp. 016555152093438
Author(s):  
Jose L. Martinez-Rodriguez ◽  
Ivan Lopez-Arevalo ◽  
Ana B. Rios-Alvarado
Keyword(s):  
Semantic Web ◽  
Natural Language ◽  
Information Extraction ◽  
Knowledge Base ◽  
Semantic Similarity ◽  
Similarity Measure ◽  
Real World ◽  
Semantic Similarity Measure ◽  
Web Standards ◽  
Extract Information

The Semantic Web provides guidelines for the representation of information about real-world objects (entities) and their relations (properties). This is helpful for the dissemination and consumption of information by people and applications. However, the information is mainly contained within natural language sentences, which do not have a structure or linguistic descriptions ready to be directly processed by computers. Thus, the challenge is to identify and extract the elements of information that can be represented. Hence, this article presents a strategy to extract information from sentences and its representation with Semantic Web standards. Our strategy involves Information Extraction tasks and a hybrid semantic similarity measure to get entities and relations that are later associated with individuals and properties from a Knowledge Base to create RDF triples (Subject–Predicate–Object structures). The experiments demonstrate the feasibility of our method and that it outperforms the accuracy provided by a pattern-based method from the literature.

Information Extraction for Call for Paper

2015 ◽  
Vol 6 (4) ◽  
pp. 35-49 ◽  
Author(s):  
Laurent Issertial ◽  
Hiroshi Tsuji
Keyword(s):  
Text Mining ◽  
Information Extraction ◽  
False Positive ◽  
Relevant Information ◽  
End User ◽  
Rule Based ◽  
Complex Queries ◽  
Mining Algorithm ◽  
Text Searching ◽  
Searching Method

This paper proposes a system called CFP Manager specialized on IT field and designed to ease the process of searching conference suitable to one's need. At present, the handling of CFP faces two problems: for emails, the huge quantity of CFP received can be easily skimmed through. For websites, the reviewing of some of the main CFP aggregators available online points out the lack of usable criteria. This system proposes to answer to these problems via its architecture consisting of three components: firstly an Information Extraction module extracting relevant information (as date, location, etc...) from CFP using rule based text mining algorithm. The second component enriches the now extracted data with external one from ontology models. Finally the last one displays the said data and allows the end user to perform complex queries on the CFP dataset and thus allow him to only access to CFP suitable for him. In order to validate the authors' proposal, they eventually process the well-known precision / recall metric on our information extraction component with an average of 0.95 for precision and 0.91 for recall on three different 100 CFP dataset. This paper finally discusses the validity of our approach by confronting our system for different queries with two systems already available online (WikiCFP and IEEE Conference Search) and basic text searching approach standing for searching in an email box. On a 100 CFP dataset with the wide variety of usable data and the possibility to perform complex queries we surpass basic text searching method and WikiCFP by not returning the false positive usually returned by them and find a result close to the IEEE system.

scholarly journals P089: Does the use of ultrasound improve diagnosis during simulated trauma scenarios?

CJEM ◽  
2017 ◽  
Vol 19 (S1) ◽  
pp. S108
Author(s):  
D. McLean ◽  
L. Hewitson ◽  
D. Lewis ◽  
J. Fraser ◽  
J. Mekwan ◽  
...  
Keyword(s):  
Differential Diagnosis ◽  
Diagnostic Performance ◽  
Statistical Tests ◽  
Correct Diagnosis ◽  
Primary Diagnosis ◽  
Point Of Care Ultrasound ◽  
Trauma Patients ◽  
Diagnostic Capabilities ◽  
Attending Physicians ◽  
The Impact

Introduction: Point of care ultrasound (US) is a key adjunct in the management of trauma patients, in the form of the extended focused assessment with sonography in trauma (E-FAST) scan. This study assessed the impact of adding an edus2 ultrasound simulator on the diagnostic capabilities of resident and attending physicians participating in simulated trauma scenarios. Methods: 12 residents and 20 attending physicians participated in 114 trauma simulations utilizing a Laerdal 3G mannequin. Participants generated a ranked differential diagnosis list after a standard assessment, and again after completing a simulated US scan for each scenario. We compared reports to determine if US improved diagnostic performance over a physical exam alone. Standard statistical tests (χ2 and Student t tests) were performed. The research team was independent of the edus2 designers. Results: Primary diagnosis improved significantly from 53 (46%) to 97 (85%) correct diagnoses with the addition of simulated US (χ2=37.7, 1df; p=<0.0001). Of the 61 scenarios where an incorrect top ranked diagnosis was given, 51 (84%) improved following US. Participants were assigned a score from 1 to 5 based on where the correct diagnosis was ranked, with a 5 indicating a correct primary diagnosis. Median scores significantly increased from 3.8 (IQR 3, 4.9) to 5 (IQR 4.7, 5; W=219, p<0.0001).Participants were significantly more confident in their diagnoses after using the US simulator, as shown by the increase in their mean confidence in the correct diagnosis from 53.1% (SD 22.8) to 83.5% (SD 19.1; t=9.0; p<0.0001)Additionally, participants significantly narrowed their differential diagnosis lists from an initial medium count of 3.5 (IQR 2.9, 4.4) possible diagnoses to 2.4 (IQR 1.9, 3; W=-378, p<0.0001) following US. The performance of residents was compared to that of attending physicians for each of the above analyses. No differences in performance were detected. Conclusion: This study showed that the addition of ultrasound to simulated trauma scenarios improved the diagnostic capabilities of resident and attending physicians. Specifically, participants improved in diagnostic accuracy, diagnostic confidence, and diagnostic precision. Additionally, we have shown that the edus2 simulator can be integrated into high fidelity simulation in a way that improves diagnostic performance.

Thoracic duct cysts: A rare differential diagnosis

2005 ◽  
Vol 132 (2) ◽  
pp. 330-333 ◽  
Author(s):  
Frank Gottwald ◽  
Heinrich Iro ◽  
Carsten Finke ◽  
Johannes Zenk
Keyword(s):  
Differential Diagnosis ◽  
Thoracic Duct ◽  
Jugular Vein ◽  
Primary Diagnosis ◽  
Imaging Method ◽  
Left Internal Jugular Vein ◽  
Duct Cyst ◽  
X Ray ◽  
Adjacent Structures ◽  
Rare Differential Diagnosis

OBJECTIVE: Cysts of the thoracic duct located in the supraclavicular region are uncommon. To date only 12 cases in this topographic area have been described in the literature. Between 1998 and 2002, 5 patients presented to our department with the primary symptom of a palpable soft left-supracavicular swelling that could be displaced relative to adjacent structures. SETTING: In each case, sonography showed a hypoechogenic, almost echo-free, distinctly outlined polycyclic structure with distal echo enhancement at the junction of the left internal jugular vein and the subclavian vein. All 5 patients underwent surgery, the cysts were extirpated, and the numerous communicating lymph vessels localized and meticulously ligated. Pathohistologic analysis of the milky, yellowish fluid obtained by intraoperative puncture confirmed the initial suspicion of a thoracic duct cyst in all patients. CONCLUSION: In the case of left supraclavicular masses, the rare differential diagnosis of a thoracic duct cyst must be considered as a possibility. Sonography as the imaging method of choice is sufficient for primary diagnosis. In addition, a thorax x-ray should be performed in order to exclude an intrathoracic involvement. Surgical extirpation marks the therapy of choice in treating such cysts.

scholarly journals Analysis of E-mental Health Research: Mapping the Relationship between Information Technology and Mental Healthcare

2021 ◽  
Author(s):  
tatsawan timakum ◽  
Min Song ◽  
Qing Xie
Keyword(s):  
Mental Health ◽  
Information Technology ◽  
Text Mining ◽  
Information Extraction ◽  
Health Research ◽  
Health Care Services ◽  
Text Message ◽  
Mental Healthcare ◽  
Mental Health Research ◽  
Mental Wellbeing

Abstract Background: E-mentalhealthcare is the convergence of digital technologies with mental health services. It has beendevelopedto fill a gap in healthcare for people who need mental wellbeing support and may never otherwise receive psychological treatment.This study aimed to apply text mining techniques to analyze the huge data of e-mental health researches and to report on research clusters and trends as well as the co-occurrence of biomedical and the use of information technology in this field.Methods: The e-mentalhealth research data was obtainedfrom 3,663 bibliographicrecords from Web of Science (WoS)and 3,172 full-text articlesfrom PubMed Central (PMC). The text mining techniques utilized for this study includedbibliometric analysis, information extraction, and visualization.Results: The e-mental health research topic trendsprimarily involvede-health care services and medical informatics research. The clusters of research comprise 16 clusters, which refer to mental sickness, ehealth, diseases, IT, and self-management. Based onthe information extraction analysis, in the biomedical domain, a “depression” entity was frequently detected and it pairs with other entities in the network with a betweenness centrality weighted at 0.046869 (eg. depression-online, depression-diabetes, depression-measure, and depression-mobile).The IT entity-relations of “mobile” were the most frequently found(weighted at 0.043466). The top pairs are related to depression, mobile health, and text message.Conclusions: E-mental health research trends focused on disease related-depression and using IT for treatment and prevention, primarily via online and mobile devices. Producing AI and machine learning are also being studied for e-mental healthcare. The results illustrate that physical sickness is likely to cause a mental health problem and identify the IT that was applied to help manage and mitigate mental health impacts.

Sign in / Sign up

Export Citation Format

Share Document