Effect of computed tomography display window and image plane on diagnostic certainty for characteristics of dysplastic elbow joints in dogs

To test the effects of computed tomography (CT) image plane and window settings on diagnostic certainty for CT characteristics associated with dysplastic elbow joints (elbow joint dysplasia) in dogs and to provide optimal display guidelines for these CT characteristics. CT images of 50 dysplastic elbow joints from 49 lame dogs and 10 elbow joints from 5 sound dogs. CT image data were obtained in transverse, sagittal, and dorsal planes. Each plane was examined by use of 3 Hounsfield unit (HU) window settings. Two veterinary radiologists independently evaluated sets of CT images for evidence of 7 CT characteristics. Effect of elbow joint status, image plane, and window settings on diagnostic certainty for these CT characteristics was tested by use of a visual analogue scale. Diagnostic certainty for abnormalities of the medial coronoid process (MCP) and radial incisure was highest in the transverse plane, subchondral defects or sclerosis of the trochlea humeri was highest in the dorsal plane, and joint incongruity was highest in the sagittal plane. Certainty for hypoattenuating subchondral defects or fissures was highest at 2,500 or 3,500 HUs, whereas certainty for subchondral sclerosis was highest at 1,500 HUs and lowest at 3,500 HUs. Diagnostic certainty for CT characteristics of elbow joint dysplasia in dogs was affected by image display variables. Diagnostic certainty for altered subchondral bone density was primarily influenced by window settings, whereas structural MCP abnormalities and joint incongruity were influenced most by image plane.