Three-dimensional volumetric ultrasonography. Does it improve reliabililty of musculoskeletal ultrasound?

Objective. To compare the interobserver reliability of three-dimensional (3D) volumetric ultrasonography (US) and 2D real-time US in detecting inflammatory and destructive changes in rheumatoid arthritis (RA) wrist and hand. Methods. Two RA patients were selected by a rheumatologist who performed independently a grey-scale and power Doppler (PD) volumetric acquisition at three anatomic sites in their more symptomatic wrist/hand using two identical scanners equipped with 3D volumetric probe. Twelve rheumatologists expert in MSUS were randomly assigned to a US scanner and a patient. In the first part of the study, each group of experts blindly, independently, and consecutively performed a 2D real-time grey-scale and PD US investigation of inflammatory changes and/or bone erosions at the three anatomic sites. In the second part of the study, each group of investigators blindly evaluated the same pathologic changes in the 6 volumes from the patient not scanned by them. Results. The kappa values were higher for 3D volumetric US than for 2D US in the detection of synovitis/tenosynovitis (0.41 vs. 0.37) and PD signal (0.82 vs 0.45) and in the PD signal grading (0.81 vs. 0.55). Conclusion. 3D volumetric US may improve the interobserver reliability in RA multicentre studies.