Diagnostic Feasibility of Tomosynthesis for Evaluation of Subacromial Impingement

Abstract

OBJECTIVE To correlate acromio-humeral distance (AHD) measured using tomosynthesis and rotator cuff (RC) pathology and to assess the diagnostic feasibility of tomosynthesis for evaluation of subacromial impingement.

MATERIALS AND METHODS A retrospective review of 63 patients with clinically suspected subacromial impingement (31 men, 32 women; mean age, 60.0 years; range, 40-72 years) was carried out at a single institution. Two musculoskeletal radiologists independently measured the following quantitative data: AHD on plain radiography and AHD at three compartments (anterior, middle, and posterior) using tomosynthesis, computed tomography (CT) arthrography, or magnetic resonance (MR) arthrography. To investigate the association between AHD and RC pathology, we reviewed the arthroscopic operation record of each patient for a reference standard. Independent t-test, Pearson correlation, two-way analysis of variance (ANOVA), and linear regression were used for statistical analysis of differences among the three modalities.

RESULTS The size of RC tendon tear (RCT) showed a significant inverse correlation with posterior tomosynthetic AHD and all CT or MR arthrographic AHDs (p<0.05). The size of RCT in full-thickness tears showed a significant inverse correlation with middle and posterior tomosynthetic AHDs and all CT or MR arthrographic AHDs (p<0.05). The type of RCT showed a significant inverse correlation with all CT or MR arthrographic AHDs (p<0.005). After ANOVA, middle tomosynthetic AHD retained a significant association with type of RCT (p=0.042), and posterior tomosynthetic AHD retained significance for size of RCT in full-thickness tear (p=0.024). After linear regression, a significant association remained between middle and posterior tomosynthetic AHDs and size of RCT in full-thickness tear (r=0.345, p=0.025; r=0.365, p=0.017) and between posterior tomosynthetic AHD and size of RCT (r=0.312, p=0.047). Intraobserver and interobserver correlation coefficients (ICCs) showed excellent agreement (ICC= 0.896-0.983) during AHD measurement.

CONCLUSION AHD measurement using tomosynthesis is practical and feasible. Middle and posterior AHDs on tomosynthesis may be practical for predicting the size of RCT in subacromial impingement.

Keywords: tomosynthesis, shoulder impingement, subacromial space, rotator cuff