Quantitative perfusion histogram parameters of dynamic contrast-enhanced MRI to identify different pathological types of uterine leiomyoma / 浙江大学学报·医学版

ABSTRACT

To explore the value of quantitative perfusion histogram parameters of dynamic contrast-enhanced magnetic resonance imaging （DCE-MRI） in pathological classification of uterine leiomyoma and its correlation with Ki-67 protein expression. Thirty five patients with uterine leiomyoma confirmed by operation and pathology at Shaoxing People's Hospital from October 2015 to September 2017 were analyzed retrospectively，including 15 cases of ordinary type，8 cases of cellular type and 12 cases of degenerative type. All patients were examined by pelvic DCE-MRI before operation，and the histogram parameters （median，mean，skewness，kurtosis，energy，entropy） of various quantitative perfusion parameters，including volume transport constant （K），rate constant （K），extravascular extracellular space distribute volume per unit tissue volume （V），blood plasma volume per unit volume of tissue （V） were calculated，and the efficacy of different parameters in pathological classification of uterine leiomyoma was evaluated by ROC curve. The expression of Ki-67 protein in uterine leiomyoma was detected by immunohistochemical method，and the correlation between histogram parameters and Ki-67 protein expression was analyzed by Pearson and Spearman correlation analysis. The median and mean values of K，K，V and V in the cellular group were higher than those in the degenerative group and the ordinary group（<0.05 or <0.01），while the skewness of V，the skewness and kurtosis of K in the cellular group were lower than those in the ordinary group （all <0.05）. The entropy of K in the cellular group was higher than that in the degenerative group and the ordinary group （all < 0.05）. The entropy of V in the cellular group was higher than that in the ordinary group （<0.01）. The median，mean，skewness of K，median and mean of K，median and mean of V，median，mean，energy and entropy of V were correlated with Ki-67 expression（all <0.05）. The results of ROC curve analysis showed that the median threshold of K was 0.994/min，the sensitivity and specificity for the diagnosis of cellular uterine leiomyoma were 100.0% and 77.8% respectively，and the area under the ROC curve was 0.949. When the mean threshold of K was 1.170/min，the sensitivity and specificity for diagnosing cellular uterine leiomyoma were 100.0% and 77.8% respectively，and the area under the ROC curve was 0.958. The area under the ROC curve of K （entropy），K （median，mean），V （median，mean，entropy） in the diagnosis of cellular uterine leiomyoma were 0.755-0.907.DCE-MRI quantitative perfusion histogram parameters have high diagnostic value in differentiating pathological types of uterine leiomyoma，especially for cellular uterine leiomyoma.