Document Detail


Automatic classification of left ventricular wall segments in small animal ultrasound imaging.
MedLine Citation:
PMID:  25053013     Owner:  NLM     Status:  Publisher    
Abstract/OtherAbstract:
Multiple statistics show that heart diseases are one of the main causes of mortality in our highly developed societies today. These diseases lead to a change of the physiology of the heart, which gives useful information about characteristic and severity of the defect. A fast and reliable diagnosis is the base for successful therapy. As a first step towards recognition of such heart remodeling processes, this work proposes a fully automatic processing pipeline for regional classification of the left ventricular wall in ultrasound images of small animals. The pipeline is based on state-of-the-art methods from computer vision and pattern classification. The myocardial wall is segmented and its motion is estimated. A feature extraction using the segmented data is realized to automatically classify the image regions into normal and abnormal myocardial tissue. The performance of the proposed pipeline is evaluated and a comparison of common classification algorithms on ultrasound data of living mice before and after artificially induced myocardial infarction is given. It is shown that the results of this work, reaching a maximum accuracy of 91.46%, are an encouraging base for further investigation.
Authors:
Kathrin Ungru; Daniel Tenbrinck; Xiaoyi Jiang; Jörg Stypmann
Related Documents :
24490083 - Histological quantification of angiogenesis after focal cerebral infarction: a systemat...
3969513 - Disturbances of the heart structure and function in chronic hemolytic anemia, their com...
9587163 - Beneficial effect of captopril against ischaemia-reperfusion injury in isolated guinea ...
Publication Detail:
Type:  JOURNAL ARTICLE     Date:  2014-6-30
Journal Detail:
Title:  Computer methods and programs in biomedicine     Volume:  -     ISSN:  1872-7565     ISO Abbreviation:  Comput Methods Programs Biomed     Publication Date:  2014 Jun 
Date Detail:
Created Date:  2014-7-23     Completed Date:  -     Revised Date:  -    
Medline Journal Info:
Nlm Unique ID:  8506513     Medline TA:  Comput Methods Programs Biomed     Country:  -    
Other Details:
Languages:  ENG     Pagination:  -     Citation Subset:  -    
Copyright Information:
Copyright © 2014 Elsevier Ireland Ltd. All rights reserved.
Export Citation:
APA/MLA Format     Download EndNote     Download BibTex
MeSH Terms
Descriptor/Qualifier:

From MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine


Previous Document:  Physical Activity and Physical Self-Concept in Youth: Systematic Review and Meta-Analysis.
Next Document:  Midterm outcomes of protection for upper urinary tract function by augmentation enterocystoplasty in...