基于经皮肩部穿刺超声造影诊断肩袖撕裂的一致性

目的 观察基于经皮肩部穿刺超声造影（CEUS）诊断各型肩袖撕裂（RCT）的一致性。方法 回顾性分析94例临床疑诊RCT患者的肩部二维超声、CEUS单一经皮超声引导肩峰下滑囊超声造影（PUSB）或联合经皮超声引导肌腱病损区超声造影（PUTL）]及关节镜检查资料，分别由观察者1（主任医师）、观察者2（主治医师）、观察者3（住院医师）基于肩部穿刺CEUS评估有无RCT及其类型，观察其诊断一致性及效能。结果 94例中，关节镜检出17例较大全层撕裂（lF）、14例小全层撕裂（sF）、18例滑囊侧部分撕裂（bP）、23例关节侧部分撕裂（aP）及22例无撕裂（NT）。观察者间对全部病例的评估结果的一致性为较高或高（Kappa=0.67~0.95）。针对全层撕裂（FT）、部分撕裂（PT）、lF、sF、bP、aP和NT，观察者1 vs. 2、2 vs. 3的一致性均高（Kappa=0.81~0.95）、观察者1 vs. 3均较高（Kappa=0.71~0.80）。观察者诊断RCT的准确率为87.23%~96.81%。针对RCT，观察者3诊断bP的敏感度为77.78%（14/18），3名观察者诊断的敏感度及特异度均>80.00%，对lF及NT的诊断敏感度均为100%（17/17）及90.91%（20/22）。结论 经皮肩部穿刺CEUS诊断各型RCT的一致性及敏感度、特异度均较高。

Consistency of diagnosing rotator cuff tear based on contrast-enhanced ultrasound through percutaneous shoulder puncture

Objective To investigate the consistency of diagnosing rotator cuff tear (RCT) by different experienced observers based on contrast-enhanced ultrasound (CEUS) through percutaneous shoulder puncture. Methods Data of two-dimensional ultrasound of shoulder, CEUS (percutaneous ultrasound-guided subacromial bursographyPUSB] alone or combined with percutaneous ultrasound-guided tendon lesionographyPUTL]) and arthroscopy of 94 patients with clinically suspected RCT were retrospectively analyzed. The classification and subtypes of RCT were evaluated by observers 1 (chief physician), 2 (attending physician) and 3 (resident physician) based on CEUS, and the diagnostic consistency and efficacy were analyzed. Results Shoulder arthroscopy showed 17 cases of larger full-thickness tears (lF), 14 of small full-thickness tears (sF), 18 of bursal-side partial-thickness tears (bP), 23 of articular-side partial-thickness tears (aP) and 22 cases without tear (NT). For all patients and each subtype of RCT, the consistencies of observers based on percutaneous shoulder puncture CEUS were fairly high or high (Kappa=0.67-0.95). For full-thickness tear (FT), partial-thickness tear (PT), lF, sF, bP, aP, and NT, the consistency of observer 1 vs. 2 and 2 vs. 3 were all high (Kappa=0.81-0.95), while of 1 vs. 3 were all fairly high (Kappa=0.71-0.80). For all patients, the diagnostic accuracy of RCT ranged from 87.23% to 96.81%. For each type of RCT, the sensitivity of observer 3 for diagnosing bP was 77.78% (14/18), the sensitivity and specificity of 3 observers were all >80.00%, while the sensitivity for lF and NT was 100% (17/17) and 90.91% (20/22), respectively. Conclusion The consistency, sensitivity and specificity of diagnosing different subtypes RCT based on CEUS through percutaneous shoulder puncture were all high.