中国医疗设备2016，Vol.31Issue(2)：44-47,4.DOI:10.3969/j.issn.1674-1633.2016.02.010
基于肾动态定量分析中肾脏ROI的自动勾画研究
A Research on the Automatic ROI Delineate Method of Renal Imaging Based on Morphological Reconstruction and Otsu Thresholding Segmentation
摘要
Abstract
Objective To investigate the automatic approach to acquire the single photon emission computed tomography (SPECT) function images of kidney based on morphological reconstruction and Otsu thresholding segmentation.Methods First, we elevated the contrast ratio of the SPECT images for preprocessing. Second, we conducted morphological reconstruction, which could sharpen the edge of kidneys. Then we proceeded to identify the kidney by Otsu thresholding segmentation. Lastly, we got the boundary to achieve the segmentation to realize the automatic region-of-interest (ROI) delineation of the kidneys.Results The automatic ROI delineate method produces better reproducibility, consistency, and accuracy in renal imaging compared with manual method to delineate the kidney.Conclusion The mentioned morphological reconstruction and Otsu thresholding method have great robustness. This procedure can obtain a more rapid, accurate and stable calculation of the glomerular ifltration rate, which makes it signiifcant for clinical application.关键词
图像分割/肾小球滤过率/形态学重建/最大类间方差法Key words
image segmentation/glomerular filtration rate/morphological reconstruction/Otsu thresholding分类
信息技术与安全科学引用本文复制引用
田书畅,徐磊,蒋红兵..基于肾动态定量分析中肾脏ROI的自动勾画研究[J].中国医疗设备,2016,31(2):44-47,4.基金项目
南京市医学发展基金“青年工程”人才培养专项经费资助项目（QRX11033)。 （）
