机构地区:[1]郑州大学第三附属医院医学影像科,河南郑州450051
出 处:《中华实用诊断与治疗杂志》2025年第1期78-83,共6页Journal of Chinese Practical Diagnosis and Therapy
基 金:河南省科技发展计划项目(212102310694)。
摘 要:目的 观察乳腺肿块患者合成MRI定量参数直方图特征变化,探讨其与弥散加权成像序列表观扩散系数(ADC)联合鉴别诊断乳腺肿块性质的价值。方法 回顾性分析2022年3月—2023年6月郑州大学第三附属医院行手术治疗的63例(66个肿块)乳腺肿块患者的临床资料。63例术前均行乳腺动态增强及合成MRI扫描,记录弥散加权成像序列ADC和合成MRI定量参数直方图特征[T_(1)、T_(2)和质子密度(PD)的第10百分位数(10th)、第90百分位数(90th)、峰度、偏度、熵、平均值、四分位数间距、中位数、最大值、最小值、范围、均匀性、标准差]。根据手术组织病理结果,66个肿块中27个良性肿块(25例)为良性组,39个恶性肿块(38例)为恶性组,比较2组ADC及合成MRI定量参数直方图特征;采用多因素logistic回归分析乳腺肿块为恶性的影响因素;绘制ROC曲线,评估T_(1)-峰度、ADC鉴别诊断乳腺肿块良恶性质的效能。结果 良性组T_(1)-峰度[2.60(2.16,3.11)ms]、T_(2)-峰度[(3.26±1.10)ms]、T_(2)-偏度[(0.71±0.45)ms]、PD-峰度[(3.71±2.69)pu]均低于恶性组[3.44(2.59,5.52)ms、(5.72±4.14)ms、(1.00±0.87)ms、(4.66±2.18)pu](U=-2.837,t=-2.563~-2.145,P<0.05),PD-90th[(100.40±19.49)pu]、ADC[(1.41±0.46)10^(-3) mm^(2)s^(-1)]均高于恶性组[(93.35±17.28)pu、(1.00±0.35)10^(-3) mm^(2)s^(-1)](t=-2.511、-4.037,P均<0.05),其余T_(1)、T_(2)、PD定量参数直方图特征比较差异均无统计学意义(P>0.05)。T_(1)-峰度(OR=1.772,95%CI:1.134~2.767,P=0.012)、ADC(OR=0.042,95%CI:0.006~0.278,P=0.001)是乳腺肿块为恶性的影响因素。T_(1)-峰度、ADC分别以3.50 ms、1.09 10^(-3) mm^(2)s^(-1)为最佳截断值,诊断乳腺肿块为恶性的AUC分别为0.700(95%CI:0.582~0.812,P=0.005)、0.790(95%CI:0.674~0.884,P<0.001),二者联合诊断乳腺肿块为恶性的AUC为0.870(95%CI:0.759~0.937,P<0.001)。结论 恶性乳腺肿块T_(1)-峰度升高,ADC降低,二者联合对乳腺肿块良恶性质的诊断价值较高。Objective To observe the changes in synthetic MRI quantitative parameters histogram features in patients with breast masses,and to investigate its value to the differential diagnosis of breast masses when combined with apparent diffusion coefficient(ADC)of diffusion-weighted imaging sequence.Methods Sixty-three patients(66breast masses)underwent surgical treatment in the Third Affiliated Hospital of Zhengzhou University from March 2022to June 2023,and their clinical data were retrospectively analyzed.Before surgery,dynamic enhancement and synthetic MRI breast scanning were performed to record diffusion-weighted imaging sequence ADC and synthetic MRI quantitative parameters histogram features as 10th percentile(10th),90th percentile(90th),kurtosis,skewness,entropy,mean,interquartile,median,maximum,minimum,range,uniformity,and standard deviation of T_(1),T_(2and) proton density (PD).In 66masses,there were 27 masses in 25patients(the benign group)and 39 masses in 38patients(the malignant group)according to the histopathological results.The ADC and synthetic MRI quantitative parameters histogram features were compared between two groups.Multivariate logistic regression was used to analyze the influencing factors of malignant breast mass.ROC curves were drawn to evaluate the efficiencies of T_(1)-kurtosis and ADC on differential diagnosing benign and malignant breast masses.Results The T_(1)-kurtosis,T_2-kurtosis,T_2-skewness and PD-kurtosis were lower in the benign group[2.60(2.16,3.11)ms,(3.26±1.10)ms,(0.71±0.45)ms,(3.71±2.69)pu]than those in the malignant group[3.44(2.59,5.52)ms,(5.72±4.14)ms,(1.00±0.87)ms,(4.66±2.18)pu](U=-2.837,t=-2.563to-2.145;all P values<0.05).The PD-90th and ADC were higher in the benign group[(100.40±19.49)pu,(1.41±0.46)10^(-3 )mm^(2)s^(-1)]than those in the malignant group[(93.35±17.28)pu,(1.00±0.35)10^(-3 )mm^(2)s^(-1)](t=-2.511,-4.037;all Pvalues<0.05).There were no significant differences in the other quantitative parameters histogram features of T_(1),T_(2)and PD between two g
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...