Computer aided medical diagnosis tool to detect normal/abnormal studies in digital MR brain images

Juan Carlos Gutierrez Caceres, Christian Portugal-Zambrano, César Beltrán-Castañón

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

4 Scopus citations

Abstract

This work presents a model to support medical diagnosis through the classification of abnormality normality in medical brain images, in order to help to specialist as a previous step in the brain pathology diagnosis. Our proposal was incorporated into a content-based image retrieval system, thus we developed a useful tool for radiologists. The first step produces the features vector of MR image using Gabor Filter for the data train and test, then as second step features vector of training data are indexed into CBIR module. The third step makes the training of SVM and as four step the test dataset is classified with the SVM trained. Finally, the result of classification are presented with a set of similar images product of a KNN query. This model was implemented as a software tool with graphical interface. We obtained 94.12% of correct classification. Our medical image dataset is composed of 187 MRI images collected from a medical diagnosis company and selected by medical specialist. The result shows that the proposed model is robust and effective as a software tool to aid support to medical diagnostic.

Original languageEnglish
Title of host publicationProceedings - 2014 IEEE 27th International Symposium on Computer-Based Medical Systems, CBMS 2014
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages501-502
Number of pages2
ISBN (Print)9781479944354
DOIs
StatePublished - 2014
Event27th IEEE International Symposium on Computer-Based Medical Systems, CBMS 2014 - New York, NY, United States
Duration: 27 May 201429 May 2014

Publication series

NameProceedings - IEEE Symposium on Computer-Based Medical Systems
ISSN (Print)1063-7125

Conference

Conference27th IEEE International Symposium on Computer-Based Medical Systems, CBMS 2014
Country/TerritoryUnited States
CityNew York, NY
Period27/05/1429/05/14

Keywords

  • cbir
  • computer aided diagnosis
  • pattern recognition
  • svm

Fingerprint

Dive into the research topics of 'Computer aided medical diagnosis tool to detect normal/abnormal studies in digital MR brain images'. Together they form a unique fingerprint.

Cite this