S Liu6, H Shi1, C Ji1, H Zheng1, X Pan1, W Guan2, L Chen3, Y Sun4, L Tang4, Y Guan5, W Li5, Y Ge5, J He6, S Liu6, Z Zhou7. 1. Department of Radiology, Nanjing Drum Tower Hospital, The Affiliated Hospital of Nanjing University Medical School, Nanjing, 210008, China. 2. Department of Gastrointestinal Surgery, Nanjing Drum Tower Hospital, The Affiliated Hospital of Nanjing University Medical School, Nanjing, 210008, China. 3. Department of Pathology, Nanjing Drum Tower Hospital, The Affiliated Hospital of Nanjing University Medical School, Nanjing, 210008, China. 4. Department of Radiology, Peking University Cancer Hospital & Institute, Beijing, 100142, China. 5. School of Electronic Science and Engineering, Nanjing University, Nanjing, 210046, China. 6. Department of Radiology, Nanjing Drum Tower Hospital, The Affiliated Hospital of Nanjing University Medical School, Nanjing, 210008, China. Electronic address: hjxueren@126.com. 7. Department of Radiology, Nanjing Drum Tower Hospital, The Affiliated Hospital of Nanjing University Medical School, Nanjing, 210008, China. Electronic address: zyzhou@nju.edu.cn.
Abstract
AIM: To explore the role of computed tomography (CT) texture analysis in predicting pathologic stage of gastric cancers. MATERIALS AND METHODS: Preoperative enhanced CT images of 153 patients (112 men, 41 women) with gastric cancers were reviewed retrospectively. Regions of interest (ROIs) were manually drawn along the margin of the lesion on the section where it appeared largest on the arterial and venous CT images, which yielded texture parameters, including mean, maximum frequency, mode, skewness, kurtosis, and entropy. Correlations between texture parameters and pathological stage were analysed with Spearman's correlation test. The diagnostic performance of CT texture parameters in differentiating different stages was evaluated using receiver operating characteristic (ROC) analysis. RESULTS: Maximum frequency in the arterial phase and mean, maximum frequency, mode in the venous phase correlated positively with T stage, N stage, and overall stage (all p<0.05) of gastric cancer. Entropy in the venous phase also correlated positively with N stage (p=0.009) and overall stage (p=0.032). Skewness in the arterial phase had the highest area under the ROC curve (AUC) of 0.822 in identifying early from advanced gastric cancers. Multivariate analysis identified four parameters, including maximum frequency, skewness, entropy in the venous phase, and differentiation degree from biopsy, for predicting lymph node metastasis of gastric cancer. The multivariate model could distinguish gastric cancers with and without lymph node metastasis with an AUC of 0.892. CONCLUSION: Multiple CT texture parameters, especially those in the venous phase, correlated well with pathological stage and hold great potential in predicting lymph node metastasis of gastric cancers.
AIM: To explore the role of computed tomography (CT) texture analysis in predicting pathologic stage of gastric cancers. MATERIALS AND METHODS: Preoperative enhanced CT images of 153 patients (112 men, 41 women) with gastric cancers were reviewed retrospectively. Regions of interest (ROIs) were manually drawn along the margin of the lesion on the section where it appeared largest on the arterial and venous CT images, which yielded texture parameters, including mean, maximum frequency, mode, skewness, kurtosis, and entropy. Correlations between texture parameters and pathological stage were analysed with Spearman's correlation test. The diagnostic performance of CT texture parameters in differentiating different stages was evaluated using receiver operating characteristic (ROC) analysis. RESULTS: Maximum frequency in the arterial phase and mean, maximum frequency, mode in the venous phase correlated positively with T stage, N stage, and overall stage (all p<0.05) of gastric cancer. Entropy in the venous phase also correlated positively with N stage (p=0.009) and overall stage (p=0.032). Skewness in the arterial phase had the highest area under the ROC curve (AUC) of 0.822 in identifying early from advanced gastric cancers. Multivariate analysis identified four parameters, including maximum frequency, skewness, entropy in the venous phase, and differentiation degree from biopsy, for predicting lymph node metastasis of gastric cancer. The multivariate model could distinguish gastric cancers with and without lymph node metastasis with an AUC of 0.892. CONCLUSION: Multiple CT texture parameters, especially those in the venous phase, correlated well with pathological stage and hold great potential in predicting lymph node metastasis of gastric cancers.
Authors: Remy Klaassen; Ruben T H M Larue; Banafsche Mearadji; Stephanie O van der Woude; Jaap Stoker; Philippe Lambin; Hanneke W M van Laarhoven Journal: PLoS One Date: 2018-11-15 Impact factor: 3.240