Abstract Archives of the RSNA, 2011
LL-INS-WE9A
Text Dictionary Mapping in Radiology Reports: Seamless Association between Online Medical Dictionaries and Radiology Report Findings
Scientific Informal (Poster) Presentations
Presented on November 30, 2011
Presented as part of LL-INS-WE: Informatics
Supriya Gupta MBBS, Presenter: Nothing to Disclose
Thomas J. Schultz BS, Abstract Co-Author: Nothing to Disclose
Keith J. Dreyer DO, PhD, Abstract Co-Author: Medical Advisor, General Electric Company
Medical Advisor, Siemens AG
Medical Advisor, Nuance Communications, Inc
Medical Advisor, Carestream Health, Inc
Medical Advisor, Vital Images, Inc
Medical Advisor, Amirsys, Inc
Medical Advisor, Life Image Inc
Medical Advisor, McKesson Corporation
Medical Advisor, Merge Healthcare
The concept-mapping ensures that comprehension will be more reliable, accurate, and complete. This direct linkage enhances report clarity and can be enhanced to map to clinical signs and symptoms within the medical record of the patient.
Semantic, or concept-based, indexing allows users to search for information using medical terms and concepts. Text maps depict important concepts in a selection of text and show how they connect structurally. They can help develop comprehension of radiology diagnosis before, during, and/or after reading. As a teaching strategy, this can be used to train residents and medical students to understand report terms and facilitate review by attending.
We propose implementation of software to enable markup of terms in radiology text reports which can be linked to various content sources. The software is jointly developed with Agfa and integrates into Xero enterprise viewer utilizing an AJAX call to Xero server. The impression is identified and extracted from report. The text is broken down into individual concepts. Concepts are processed through noise reduction parser to remove “non-included” and duplicated terms. Content calls are generated for remaining “Important Terms” integrated into URL or local content calls. Content calls are formatted into JSON objects and returned to the calling application so that final terms can be used for content searches. Possible content sources include MedicineNet.com, Merriam-Webster, ICD9, Teaching Files (MIRC, MedPix) and Google Images.
Concept-based indices often possess taxonomical structure which enable applications to use term hierarchy to generalize the search to include lexical variants, abbreviations and synonyms. Assuming sentences are a single concept and ignoring loose associations is a limitation. This visual mode of presenting information makes it easily comprehensible in a short time. Physicians can skim through a report for a quick understanding of entire exam due to highlighting of key points making it useful for reviews or second opinions. It can also help radiologists to understand the possible indirect results of a finding since cross-links may show relation between concepts.
Gupta, S,
Schultz, T,
Dreyer, K,
Text Dictionary Mapping in Radiology Reports: Seamless Association between Online Medical Dictionaries and Radiology Report Findings. Radiological Society of North America 2011 Scientific Assembly and Annual Meeting, November 26 - December 2, 2011 ,Chicago IL.
http://archive.rsna.org/2011/11010188.html