Share Email Print
cover

Journal of Biomedical Optics • Open Access

Automated identification of abnormal metaphase chromosome cells for the detection of chronic myeloid leukemia using microscopic images
Author(s): Xingwei Wang; Bin Zheng; Shibo Li; John J. Mulvihill; Xiaodong Chen; Hong Liu

Paper Abstract

Karyotyping is an important process to classify chromosomes into standard classes and the results are routinely used by the clinicians to diagnose cancers and genetic diseases. However, visual karyotyping using microscopic images is time-consuming and tedious, which reduces the diagnostic efficiency and accuracy. Although many efforts have been made to develop computerized schemes for automated karyotyping, no schemes can get be performed without substantial human intervention. Instead of developing a method to classify all chromosome classes, we develop an automatic scheme to detect abnormal metaphase cells by identifying a specific class of chromosomes (class 22) and prescreen for suspicious chronic myeloid leukemia (CML). The scheme includes three steps: (1) iteratively segment randomly distributed individual chromosomes, (2) process segmented chromosomes and compute image features to identify the candidates, and (3) apply an adaptive matching template to identify chromosomes of class 22. An image data set of 451 metaphase cells extracted from bone marrow specimens of 30 positive and 30 negative cases for CML is selected to test the scheme's performance. The overall case-based classification accuracy is 93.3% (100% sensitivity and 86.7% specificity). The results demonstrate the feasibility of applying an automated scheme to detect or prescreen the suspicious cancer cases.

Paper Details

Date Published: 1 July 2010
PDF: 12 pages
J. Biomed. Opt. 15(4) 046026 doi: 10.1117/1.3476336
Published in: Journal of Biomedical Optics Volume 15, Issue 4
Show Author Affiliations
Xingwei Wang, Univ. of Pittsburgh Medical Ctr. (United States)
Bin Zheng, Univ. of Pittsburgh Medical Ctr. (United States)
Shibo Li, Oklahoma Medical Research Foundation (United States)
John J. Mulvihill, The Univ. of Oklahoma Health Sciences Ctr. (United States)
Xiaodong Chen, The Univ. of Oklahoma Bioengineering Ctr. (United States)
Hong Liu, The Univ. of Oklahoma (United States)


© SPIE. Terms of Use
Back to Top