【摘要】：目的:探討磁共振擴散峰度成像(DKI)所得參數(shù)與胃癌侵襲性的潛在相關(guān)性。材料與方法:該前瞻性研究共納入了 49例胃癌患者。所有患者在手術(shù)之前都接受了磁共振檢查。每位患者的磁共振圖像都導入工作站軟件IDL 6.3進行分析。兩名放射科醫(yī)師在顯示腫瘤最大病灶的DKI圖像上手動勾畫感興趣區(qū)(ROI),由軟件自動計算出三個DKI參數(shù):表觀擴散系數(shù)(ADC)、校正后擴散系數(shù)(D)和擴散峰度系數(shù)(K)。統(tǒng)計學分析使用SPSS22.0,雙側(cè)P值0.05表示有統(tǒng)計學意義。Shapiro-Wilk檢驗用于分析DKI參數(shù)在各個亞組中是否符合正態(tài)性分布,Spearman相關(guān)性檢驗用于分析DKI參數(shù)與胃癌T分期、N分期、總分期、分化程度及Lauren分型的相關(guān)性,不同N分期及Lauren分型的胃癌病灶的DKI參數(shù)的差異性應用單因素方差分析,不同T分期(T1-2vsT3-4)、總分期(Ⅰ-Ⅱ vs Ⅲ-Ⅳ)、分化程度(低/中低vs高/中)、淋巴結(jié)轉(zhuǎn)移狀態(tài)(有vs無)、脈管癌栓(有vs無)、神經(jīng)侵犯(有vs無)的胃癌病灶的DKI參數(shù)的差異性應用獨立樣本t檢驗分析。受試者工作特征曲線用于分析DKI參數(shù)鑒別胃癌不同病理特征的診斷效能。結(jié)果:ADC值與胃癌T分期、分化程度及Lauren分型呈顯著負相關(guān),K值與胃癌T分期、總分期、分化程度及Lauren分型呈顯著正相關(guān),D值均未表現(xiàn)出顯著相關(guān)性。在胃癌各個病理特征的鑒別診斷上,三個DKI參數(shù)各有其優(yōu)勢。ADC、D和K值在T1-2期與T3-4期胃癌的鑒別中有較好的表現(xiàn)(曲線下面積AUC = 0.798、0.745和0.804),ADC和D值有助于鑒別胃癌有無淋巴結(jié)轉(zhuǎn)移(AUC=0.741和0.729),ADC和K值有助于鑒別高/中分化與低/中低分化(AUC = 0.732和0.794)、腸型與混合/彌漫型(AUC = 0.678和0.755)的胃癌。而在胃癌有無脈管及神經(jīng)侵犯的評估上,三個DKI參數(shù)均未表現(xiàn)出顯著性差異。結(jié)論:磁共振擴散峰度成像有助于術(shù)前評估胃癌侵襲性,尤其是鑒別胃癌不同T分期、分化程度及Lauren分型,以及預測胃癌患者有無淋巴結(jié)轉(zhuǎn)移。
[Abstract]:Objective: to investigate the potential correlation between (DKI) parameters and invasion of gastric cancer. Materials and methods: the prospective study included 49 patients with gastric cancer. All patients underwent magnetic resonance imaging prior to surgery. The MRI images of each patient were imported into the workstation software IDL 6.3 for analysis. Two radiologists manually delineate the region of interest (ROI),) on DKI images showing the largest lesions of the tumor. The software automatically calculates three DKI parameters: apparent diffusion coefficient (ADC),). Corrected diffusion coefficient (D) and diffusion kurtosis coefficient (K). Shapiro-Wilk test was used to analyze the normal distribution of DKI parameters in each subgroup, and Spearman correlation test was used to analyze the relationship between DKI parameters and T stage of gastric cancer. The correlation between N stage, total stage, differentiation degree and Lauren classification, the difference of DKI parameters in different N staging and Lauren classification of gastric cancer was analyzed by univariate ANOVA, different T staging (T1-2vsT3-4). Total stage (鈪，

本文編號：2311180