For the evaluation of the imaging properties of medical radiographic systems there are well-established standards for measuring techniques available, e.g. ISO 9236 for the measurement of the H/D curve. However, such measuring techniques require sophisticated equipment which is not available in a clinical environment. For a clinical routine of image evaluation, techniques like contrast-detail diagrams or the visual inspection of radiographs of grid pattern with varying contrast and spatial resolution are very common. The disadvantage of these techniques is that the corresponding results are very hard to be transferred to real patient images. Therefore, observer studies on the detection of certain pathology are commonly used to e.g. investigate the influence of different radiographic techniques on diagnostic image quality. As it is very difficult to find a sufficient number of patients with real pathologies for such studies, the pathologies are often simulated by fixing e.g. aluminium disks or other nodule-like objects to healthy volunteers when the radiograph is produced. This approach is relatively simple and rather successful in chest imaging. For lumbar spine images, however, the situation is different because nodule-like tumours cannot only consist of bony material which is increasing the X-ray absorption, but tumours can also destroy the bone material resulting in an increased transparency of the corresponding anatomical region. Such a behaviour is extremely hard to be simulated by fixing an external object to the patient. However, it can be easily simulated in a computer and applied to digital radiographic data. The current paper presents a computer model for the simulation of nodules in lumbar spine images. The model has been applied within a CEC founded research project, which was investigating the influence of MTF and noise power spectra on the diagnostic quality of radiographs of the lumbar spine.
|Journal||Radiation Protection Dosimetry|
|Publication status||Published - 2000|
Subject classification (UKÄ)
- Radiology, Nuclear Medicine and Medical Imaging