Low-dose CT imaging of radio-opaque markers for assessing human rotator cuff repair: accuracy, repeatability and the effect of arm position.

Previous studies have used radiostereometric analysis (RSA) to assess the integrity and mechanical properties of repaired tendons and ligament grafts. A conceptually similar approach is to use CT imaging to measure the 3D position and distance between implanted markers. The purpose of this study was to quantify the accuracy and repeatability of measuring the position and distance between metallic markers placed in the rotator cuff using low-dose CT imaging. We also investigated the effect of repeated or variable positions of the arm on position and distance measures. Six human patients had undergone rotator cuff repair and placement of tantalum beads in the rotator cuff at least one year prior to participating in this study. On a single day each patient underwent nine low-dose CT scans in seven unique arm positions. CT scans were analyzed to assess bias, precision and RMS error of the measurement technique. The effect of repeated or variable positions of the arm on the 3D position of the beads and the distance between these beads and suture anchors in the humeral head were also assessed. Results showed the CT imaging method is accurate and repeatable to within 0.7 mm. Further, measures of bead position and anchor-to-bead distance are influenced by arm position and location of the bead within the rotator cuff. Beads located in the posterior rotator cuff moved medially as much as 20 mm in abduction or external rotation. When clinically relevant CT arm positions such as the hand on umbilicus or at side were repeated, bead position varied less than 4 mm in any anatomic direction and anchor-to-bead distance varied +2.8 to -1.6 mm (RMS 1.3 mm). We conclude that a range of ± 3 mm is a conservative estimate of the uncertainty in anchor-to-bead distance for patients repeatedly scanned in clinically-relevant arm positions.