Page 126 - Imaging of Osteoarthritis and Rheumatoid Arthritis in Hand Joints
P. 126

                                Chapter 7
happen during acquisition of images, detection of abnormalities on the images, and interpretation of these abnormalities. In all studies presented from chapter 3-6 we tried to minimize this bias. Systematical difference in image acquisition was minimized in all studies by using standardized imaging protocols, except in chapter 3 which was a retrospective study. In this study, different views were available per examination. We did not have enough data to subdivide the data for the used radiological views, but one can expect that the difference in accuracy between CT and CR can diminish if more specialized views are used for depiction of the CMC1 and STT joint, like the Bett’s view.
In chapter 6, ultrasound was used in which purposely the acquisition of images was standardized for only one of the readers. There was a significant difference between the results of those readers, but we cannot determine if this could also be partly because of systematical difference in detection and interpretation. To minimize errors in both detection and interpretation, we employed multiple readers to reduce variability in chapters 3-5, either by using a consensus result, or by using the averages of the readers. To minimize errors in interpretation, scoring systems were used in chapter 3-6 which are specifically created to increase consistency of results within and between studies. Overall, information bias will be least present in the chapters about direct cartilage imaging. Some information bias may be present, but we do not think that this would have a significant effect on our conclusions.
As a third limitation, validity of imaging methods is ideally tested against a golden standard. In hand joint research, the ideal reference standard would be histology, which is hard to obtain. In most chapters construct validity was assessed instead by comparison with other radiological imaging methods.
Future imaging methods in hand joints.
New imaging methods can be considered for future evaluation of hand joints in the future. The recently developed tomosynthesis is a new radiological imaging method, which has not been described in the previous chapters. The technique uses x-rays to acquire images from multiple angles in a limited arc. It does not fully rotate around the patient like CT. As a result, Image quality,

   124   125   126   127   128