常规MRI联合分段读出平面回波弥散加权成像在颞骨良恶性肿瘤鉴别诊断中的应用价值

目的 探讨常规MRI联合分段读出平面回波弥散加权成像(RESOLVE-DWI)鉴别颞骨良、恶性肿瘤的价值。方法 回顾性研究。纳入2014年1月-2019年12月复旦大学附属眼耳鼻喉科医院经病理学证实的96例颞骨肿瘤患者临床和影像资料。患者均在手术前2周内行颞骨常规MRI及RESOLVE-DWI检查,分别观察肿瘤大小(最大层面测量的病灶最大径)、形态、信号均匀性、边界及强化程度等图像特征,以及记录通过后处理软件在表观弥散加权成像图像上选择肿瘤实性区域手动绘制类圆形感兴趣区的表观弥散系数(ADC)。采用χ²检验和独立样本t检验比较良、恶性肿瘤的MRI信号特征及ADC值。绘制单独及联合常规MRI征像和RESOLVE-DWI测量的ADC值诊断颞骨良、恶性肿瘤的受试者工作特征曲线(ROC),并使用Delong检验比较诊断效能。结果 良性肿瘤59例,男28例、女31例,年龄(44.63±12.97)岁,肿瘤大小(2.39±0.89)cm,类圆形/浅分叶41例、不规则形18例,边界清晰52例、模糊7例,信号均匀37例、不均匀22例,轻中度强化34例、明显强化25例;恶性肿瘤37例,男18例、女19例,年龄(50.27±17.25)岁,肿瘤大小(2.58±1.16)cm,类圆形/浅分叶23例、不规则形14例,边界清晰17例、模糊20例,信号均匀27例、不均匀10例,轻中度强化9例、明显强化28例。良、恶性颞骨肿瘤间,除肿瘤边界和强化程度差异有统计学意义(P值均＜0.01)外,患者年龄、性别以及MRI图像其他特征间差异均无统计学意义(P值均>0.05)。颞骨恶性肿瘤的ADC值为(0.91 ± 0.23) ×10^-3 mm²/s,低于良性肿瘤组的 (1.09 ± 0.32) ×10^-3 mm²/s,差异有统计学意义(t=2.974, P＜0.01)。单独应用常规MRI和RESOLVE-DWI测量的ADC值诊断良、恶性颞骨肿瘤的ROC曲线下面积(AUC)分别为0.771(95%可信区间0.639~0.903)和0.727(95%可信区间0.565~0.889),其灵敏度、特异度及准确度分别为68.42%、85.71%、75.08%和63.16%、71.43%、66.35%。联合应用常规MRI征象及RESOLVE-DWI测量的ADC值诊断良、恶性颞骨肿瘤的AUC值为0.907(95%可信区间0.816~0.998),其灵敏度、特异度及准确度分别为89.47%、80.95%和86.19%。联合应用常规MRI及RESOLVE-DWI的AUC显著高于分别单独应用常规MRI和RESOLVE-DWI(Z=2.463、2.243, P值均＜0.05)。结论 常规MRI及RESOLVE-DWI在鉴别颞骨良、恶性肿瘤中具有重要价值,二者联合应用能有效提高其鉴别诊断效能。

Differential diagnosis of benign and malignant tumors of temporal bone by conventional MRI combined with readout segmentation of long variable echo-trains diffusion weighted imaging

Objective To evaluate the value of conventional MRI combined with readout segmentation of long variable echo-trains diffusion weighted imaging (RESOLVE-DWI) to differentiate the benign from the malignant tumors of the temporal bone. Methods The clinical and imaging data of 96 patients with temporal bone tumors, as confirmed by histopathology, were retrospectively analyzed in Eye and ENT Hospital of Fudan University from January 2014 to December 2019. Conventional MRI and RESOLVE-DWI were performed on all patients within 2 weeks before surgery. The following features were assessed with conventional MRI: tumor size (maximum diameters on the maximum transverse plane), shape, signal uniformity, boundary, and enhancement degree. Using post-processing software, apparent diffusion coefficient (ADC) values were obtained by drawing freehand circular regions of interest on the solid area of the tumor. The conventional MRI features and ADC values of benign and malignant tumors were compared by using the chi-square test and t-test. The receiver operating characteristic curve (ROC) of conventional MRI and ADC values alone or in combination was obtained to differentiate the benign from the malignant tumors of the temporal bone. The Delong test was used to compare the diagnostic efficiency. Results Fifty-nine patients, including 28 males and 31 females, had benign tumor of the temporal bone. They had an average age of 44.63±12.97 years old. Conventional MRI results showed that the benign tumors were 2.39±0.89 cm in size; some of these benign tumors were oval/slightly lobulated in shape (n=41), whereas others had an irregular shape (n=18), a well-defined boundary (n=52), an ill-defined boundary (n=7), a homogeneous signal (n=37), a heterogeneous signal (n=22), mild/moderate enhancement (n=34), and obvious enhancement (n=25). Thirty-seven patients, including 18 males and 19 females, had malignant tumor of the temporal bone. Their average age was 50.27±17.25 years old. Conventional MRI results showed that the malignant tumors were 2.58±1.16 cm in size; some of these malignant tumors were oval/slightly lobulated in shape (n=23), whereas others had an irregular shape (n=14), a well-defined boundary (n=17), an ill-defined boundary (n=20), a homogeneous signal (n=27), a heterogeneous signal (n=10), mild/moderate enhancement (n=9), and obvious enhancement (n=28). The tumor boundary and enhancement degree were statistically significant (all P values＜0.01) and could be used to distinguish the benign from the malignant tumors of the temporal bone. No significant differences was found for age, gender, or other features scanned by conventional MRI (all P values> 0.05). The ADC value of malignant tumor was (0.91±0.23)×10^-3 mm²/s, which was lower than that of benign tumor (1.09±0.32)×10^-3 mm²/s, and the difference was statistically significant (t=2.974, P＜0.01). The area under ROC curve (AUC) of conventional MRI in the differential diagnosis of the benign and malignant tumors of the temporal bone was 0.771 (95% CI 0.639-0.903) with sensitivity, specificity, and accuracy values of 68.42%, 85.71%, and 75.08%, respectively. The AUC of the ADC value was 0.727 (95% CI 0.565-0.889) with sensitivity, specificity, and accuracy values of 63.16%, 71.43%, and 66.35%, respectively. The AUC of the combination of conventional MRI and RESOLVE-DWI was 0.907 (95% CI 0.816-0.998) with sensitivity, specificity, and accuracy values of 89.47%, 80.95%, and 86.19%, respectively. The AUC value of the combination of conventional MRI with RESOLVE-DWI was significantly higher than the individual AUC values of these methods when used on their own (Z=2.463, 2.243, all P values＜0.05). Conclusions Conventional MRI and RESOLVE-DWI are of great value in the differential diagnosis of the benign and malignant tumors of the temporal bone. Their combination can further improve the diagnostic efficiency.