Comparison of liver volumetry on contrast‐enhanced CT images: one semiautomatic and two automatic approaches

• DOI: 10.1120/jacmp.v17i6.6485
• Publication Date: 2017-02-28T00:58:12Z
• Authors: Wei Cai; Baochun He; Yingfang Fan; Chihua Fang; Fucang Jia
• Keywords: Fully automatic

This study was to evaluate the accuracy, consistency, and efficiency of three liver volumetry methods— one interactive method, an in‐house‐developed 3D medical Image Analysis (3DMIA) system, automatic active shape model (ASM)‐based segmentation, probabilistic atlas (PA)‐guided segmentation method on clinical contrast‐enhanced CT images. Forty‐two datasets, including 27 normal 15 space‐occupying lesion patients, were retrospectively included in this study. The methods — semiautomatic 3DMIA, ASM‐based, PA‐based achieved accuracy with VD (volume difference) , 3.06% group, respectively, 4.14% respectively. However, 27.63 mins, 1.26 1.18 mins average, compared manual volumetry, which took 43.98 mins. high intraclass correlation coefficient between indicated excellent agreement volumetry. Significant differences time observed (3DMIA, ASM, PA) ( ), as well volumetries (ASM ). agreed gold standard both group group. ASM‐ have better use. PACS number(s): 87.55.‐x