应用高频超声联合数字化软件定量分析诊断2型糖尿病正中神经病变

目的研究2型糖尿病患者正中神经的回声强度及截面积，探索其诊断远端对称性多发神经病变（DSPN）的价值。 方法选取2019年6月至2020年6月在黄山市人民医院内分泌科住院的2型糖尿病患者67例，糖尿病不伴DSPN组25例（疑似组）、糖尿病伴DSPN组42例（确诊组），选取同期健康志愿者25名为对照组。应用高频超声测量非惯用手正中神经5个不同位点（MN1~5）的截面积。通过Image J软件导入超声图片，对神经内低回声区所占面积百分比进行计算。采用单因素方差分析比较3组间正中神经截面积和神经内低回声区所占面积百分比的差异，组间两两比较采用LSD-t检验；采用Pearson相关分析评价2型糖尿病患者正中神经的截面积与神经内低回声区所占面积百分比的相关性。绘制神经内低回声区所占面积百分比诊断DSPN的受试者操作特征（ROC）曲线。 结果3组间正中神经截面积、神经内低回声区所占面积百分比总体差异具有统计学意义（P＜0.05），其中确诊组各位点神经截面积均大于对照组［MN1：（9.43±1.79）mm vs（7.23±1.22）mm；MN2：（10.07±1.74）mm vs（7.49±1.30）mm；MN3：（7.61±1.23）mm vs（5.94±1.03）mm；MN4：（7.15±1.03）mm vs（5.24±0.69）mm；MN5：（13.51±2.85）mm vs（11.03±1.78）mm］，差异均具有统计学意义（t=5.843、6.701、5.816、7.597、4.110，P均＜0.001）；确诊组各位点神经低回声区所占面积百分比均大于对照组［MN1：（66.61±5.17）% vs（54.66±3.39）%；MN2：（63.50±6.46）% vs（53.91±3.09）%；MN3：（58.16±4.20）% vs（49.61±2.64）%；MN4：（54.84±4.94）% vs（45.63±4.08）%；MN5：（59.42±4.79）% vs（55.06±3.44）%］，差异均具有统计学意义（t=9.209、6.835、8.266、8.284、3.949，P均＜0.001）。以55.64%为截断值时，MN2位点处神经内低回声区所占面积百分比诊断DSPN的特异度最高为91.0%，敏感度为76.0%，曲线下面积为0.889。MN1、MN2、MN3、MN4位点处的截面积与神经内低回声区百分比存在正相关关系（r=0.442、0.554、0.387、0.463；P均＜0.01），其中MN2位点处的相关性最高。 结论2型糖尿病伴有DSPN患者的正中神经弥漫性增粗，神经内低回声区占比增大，可以在一定程度上辅助早期诊断DSPN。

Echo intensity and neuromorphology of the median nerve in patients with type 2 diabetes mellitus

ObjectiveTo investigate the echo intensity and cross-sectional area of the median nerve in patients with type 2 diabetes mellitus. MethodsSixty-seven patients with type 2 diabetes mellitus were enrolled in this study at Huangshan City People's Hospital, including 25 patients without diabetic distal symmetrical polyneuropathy (DSPN) and 42 with DSPN. Twenty-five healthy subjects were included as controls. High-frequency ultrasound was used to measure the cross-sectional area (CSA) of five different sites of the median nerve (MN) of the non-dominant hand. Image J software was used to calculate the percentage of the hypoechoic area in the median nerve. The correlation between median nerve CSA and percentage of the hypoechoic area was analyzed. ResultsThe overall differences of median nerve CSA and percentage of the hypoechoic area (HA) among the three groups were statistically significant (P＜0.05). The CSA values of patients with DSPN was significantly larger than those of healthy controls MN1: (9.43±1.79) mm vs (7.23±1.22) mm, t=5.843, P＜0.001; MN2: (10.07±1.74) mm vs (7.49±1.30) mm, t=6.701, P＜0.001; MN3: (7.61±1.23) mm vs (5.94±1.03) mm, t=5.816, P＜0.001; MN4: (7.15±1.03) mm vs (5.24±0.69) mm, t=7.597, P＜0.001; MN5: (13.51±2.85) mm vs (11.03±1.78) mm, t=4.110, P＜0.001]. The percentages of HA of patients with DSPN were significantly higher than those of healthy controls MN1: (66.61±5.17)% vs (54.66±3.39)%, t=9.209, P＜0.001; MN2: (63.50±6.46)% vs (53.91±3.09)%, t=6.835, P＜0.001; MN3: (58.16±4.20)% vs (49.61±2.64)%, t=8.266, P＜0.001; MN4: (54.84±4.94)% vs (45.63±4.08)%, t=8.284, P＜0.001; MN5: (59.42±4.79)% vs (55.06±3.44)%, t=3.949, P＜0.001]. When 55.64% was used as the cut-off value, the percentage of the HA at MN2 had the highest specificity (91.0%) for identifying DSPN, with a sensitivity of 76.0% and area under the curve of 0.889. There was a positive correlation between CSA and the percentage of the HA at different sites of the MN except the MN5 (r=0.442, 0.554, 0.387, and 0.463, respectively; P＜0.01). ConclusionPeripheral nerves are enlarged diffusely in diabetic patients with DSPN, and the percentage of the HA increases. Both of these findings are useful for the identification of DSPN at early stage.