Standardized three dimensional computerised tomography scanner reconstructions increase the accuracy of acetabular fracture classification

Evaluate the role of four standardized 3D reconstruction views in the accurate diagnosis of acetabular fractures. Thirty-five acetabular fracture cases were selected from a tertiary centre database. Fourteen reviewers with different experience level evaluated this set of images which were provided in axial 2D views and multiplanar reconstruction (MPR) without 3D views in the first two readings then the standardized 3D views were added for the subsequent two readings. The primary outcome was the accuracy of diagnosis while inter-observer reliability and reading time as well as time needed for accurate diagnosis were evaluated. Accuracy of classification was 56.7% using the 2D and 73% using the described 3D (p < 0.001). Improvement was noted in all groups even though the expert group showed the least improvement (6.7% (p = 0.04)) and the least experience group showed the most improvement (35.7% (p < 0.001)). Average time of interpretation was 96 minutes for the 2D sets and 47 minutes for the 3D sets. Finally, the adding of the four standardized views increased the inter-observer reliability in all groups compared to the standard 2D sets with MPR. Acetabular fracture diagnosis according to Letournel classification is difficult and depends greatly on the experience of the reader. The described set of 3D images yields better accuracy and renders the diagnosis more repeatable and faster. We recommend the use of these images in classifying acetabular fractures.