Evaluation of MR-derived Simulated CT-like Images and Simulated Radiographs Compared with Conventional Radiography in Patients with Suspected Shoulder Pathology

Presentation Format: Oral presentation.

Purpose or Learning Objective: To evaluate the diagnostic value of magnetic resonance (MR)-derived computed tomography (CT)-like images and simulated radiographs compared with conventional radiographs in patients with suspected shoulder pathology.

Methods or Background: We performed 3-T MR imaging of the shoulder including a three-dimensional (3D) T1-weighted gradient-echo sequence in patients with suspected rotator cuff injury. The inverted MRI image volumes were subsequently processed to achieve enhancement of the image contrast, and a cone-beam forward projection algorithm was used to obtain intensity-inverted CT-like images and simulated radiographs. Congruency of the projections and image quality was graded using a 4-point Likert scale. Two board-certified radiologists evaluated the simulated images and the conventional radiographs, performing measurements of the acromiohumeral distance (AHD) and critical shoulder angle (CSA), as well as assessing semiquantitative scores to grade degenerative changes using the Samilson-Prieto classification and the type of the acromion shape (Bigliani classification). Agreement between the modalities was calculated using Cohen κ coefficients and Wilcoxon signed rank tests.

Results or Findings: Measurements on simulated radiographs were comparable with those on conventional radiographs with substantial to almost perfect inter- and intra-rater agreements (κ = 0.69–1.00 and κ = 0.65–0.85, respectively). Diagnostic image quality of the simulated radiographs was rated “good” to “excellent.” The 3D rotation of the simulated radiographic data set was feasible, and the two-dimensional projections were graded as congruent with the conventional radiographs in 20 of 23 patients (87.0%). A substantial agreement was found between MR-derived simulated radiographs and conventional radiographs regarding diagnostically relevant features, assessed on Y and anteroposterior projections (κ = 0.84 and 0.69 for the measurement of the CSA; κ = 0.95 and 0.60 for the measurement of the AHD; κ = 0.77 and 0.77 for grading of the Samilson-Prieto classification; κ = 0.83 and 0.67 for the grading of the Bigliani classification).

Conclusion: Assessment of diagnostically relevant features of the shoulder joint using MR-derived CT-like images and simulated radiographs was reliable and showed a diagnostic quality equivalent to that of conventional radiographs. In daily practice, CT-like images and simulated radiographs could be a useful radiation-free tool that provides additional diagnostic information in the assessment of the shoulder joint.

Publication History

Article published online:
03 June 2021