Abstract—The purpose of this study is to develop a semi-automated method to measure white matter lesion (WML) volumes in four different categories, and to assess the accuracy of the method. The four-category classification of white matter lesions has the potential of providing better etiological and functional relevance than the conventional two-category classification of periventricular and deep WMLs. The methodological steps are as follows. First, the Statistical Parametric Mapping (SPM) software is used to strip off the skull and scalp, and segment the FLAIR MR images of the brain into gray matter (GM), white matter (WM), and cerebrospinal fluid (CSF). Second, such segmentation is corrected manually. Third, the white matter is partitioned into four categories. Fourth, the mean, variance of each category is computed. Then the user selects a suitable threshold that differentiates WML from normal tissues. Finally, the WML volumes for each category are calculated. Three cases which represent mild, moderate and severe WML are used to test our proposed method. The overlap percentage between our method and the gold standard is 77%, which compares favorably with the best results of 62.2% published in the literature. Based on the test results, our semi-automated method is promising to be a useful tool in quantitative measurement of white matter lesions.

Index Terms—Computer-aided decision making, medical imaging, quantitative measurement, segmentation, white matter lesion.

Yao Shieh is with Chang Gung University, 259 Wenhua 1st Road, Guishan Dist. Taoyuan City, Taiwan. He is also with Chang Gung Memorial Hospital, No.5, Fuxing St., Guishan Dist., Taoyuan City 333, Taiwan and University of California, Irvine 101 The City Drive South Orange, CA 92868 USA (email: yshieh@gmail.com).
Chien-Hung Chang is with Chang Gung University, 259 Wenhua 1st Road, Guishan Dist. Taoyuan City, Taiwan. He is also with Chang Gung Memorial Hospital, No.5, Fuxing St., Guishan Dist., Taoyuan City 333, Taiwan.
Scott Goodwin is with University of California, Irvine 101 The City Drive South Orange, CA 92868 USA.

Cite: Yao Shieh, Chien-Hung Chang, Scott Goodwin, "Quantitative Measurement of White Matter Lesions Using Semi-automated Method," International Journal of Bioscience, Biochemistry and Bioinformatics vol. 7, no. 4, pp. 238-244, 2017.