Digital and Computed Musculoskeletal Radiography. A study on spatial resolution, diagnostic accuracy, image quality and radiation dose

Sammanfattning: The aims of the study were to investigate the applications of digital radiography to musculoskeletal radiology. In the first part of the research, the diagnostic accuracy of digital images as compared to film-screen radiographs was studied and the spatial resolution requirements evaluated. Using computed radiography (CR), the pathological changes included were erosions, soft tissue swelling, periarticular osteopenia, joint space narrowing, subperiosteal resorption and intracortical resorption, and, using digitized radiographs, fractures of the scaphoid bone. The CR images were displayed as hardcopies and on interactive workstations, the digital scaphoid images solely on a workstation. There was no significant difference in the diagnostic accuracy between CR images and film-screen radiographs for all parameters, except for periarticular osteopenia where the radiographs were significantly better than CR hardcopies but equal to CR monitor images. A CR system with imaging plates of 3.33 lp/mm resolution was adequate for the diagnosis of erosions, soft tissue swelling and joint space narrowing, but for periarticular osteopenia 5 lp/mm may be needed when using hardcopies. For subperiosteal and intracortical resorption, 2x magnification CR radiography with 5 lp/mm imaging plates was sufficient. The performance in detecting subtle scaphoid fractures improved with increasing spatial resolution. Radiographs digitized with 340 µm were inferior to the original images. A pixel size of 170 µm was statistically adequate for the diagnosis of subtle scaphoid fractures but 100 - 110 µm are recommended. In the second part of the study, CR image quality in relation to radiation dose reduction was assessed. For hand images, the exposure could be lowered to 25% of the film-screen exposure and still provide adequate image quality. For hip images, the dose could be lowered to 50%. At lower exposure levels there was a progressive deterioration in CR image quality. Images with contrast enhancement were more vulnerable to dose reduction. Contrast enhancement had, however, a distinct advantage in the evaluation of soft tissues. CR offered a simple means to achieve very low radiation doses where dose reduction is imperative and a low spatial resolution is acceptable, such as in scoliosis follow-up.