| Abstract |
Purpose : The application of advanced computerized methods οn a) three - dimensional (3d) imaging of paranasal sinuses using computed tomography (CT) data, b) image segmentation from CT images in treatment planning of prostate cancer and c) splenic volume measurements from magnetic resonance (MR) imaging data. Materials and methods : CT and MR scans were transferred to a computer workstation. a) A commercially manufactured phantom was used to define the optimum scanning parameters of a 3d CT protocol of paranasal sinuses. Dose measurements were carried out using thermoluminescent dosimeters positioned in an anthropomorphic Rando phantom. Ten patients (7 with paranasal sinuses polyps, 2 with carcinoma in the same region and 1 with choanal atresia) were examined with the 3d CT protocol. 3d reconstructions were obtained using the volume rendering method on a workstation. The 3d images were evaluated on the basis of their ability to demonstrate bony abnormalities depicted on axial CT sections. b) Ten patients with prostate cancer underwent treatment planning CT prior to radiotherapy. Prostate, bladder and rectum were segmented in all CT images using the region growing technique and manual tracing. The volumes of the above organs computed with the region growing technique were compared with those from manually traced images, on a slice - by - slice basis. Measurement reproducibility of both segmentation techniques was evaluated using the data obtained from four independent observers. c) MR imaging was generated in phantoms and 16 patients with indications unrelated to splenic disease. Spleen volume measurements were performed using the random marking technique and the conventional technique of manual planimetry. Two observers independently measured splenic volume in order to evaluate the reproducibility of both volumetric techniques. Results : a) The parameters of the proposed CT protocol for 3d paranasal sinuses imaging were 120 kV, 175 mA, 1.9 sec with contiguous 1.5 mm axial slices. The use of the 3d CT protocol resulted in lower patient dose than the conventional CT examination of paranasal sinuses. The 3d images provided an ideal depiction of all bony abnormalities and anatomical variations of bony structures. b) The region growing technique was one and a half times faster than the manual tracing for prostate, bladder and rectum volume segmentation. There was no statistical difference between the slice volumes of prostate, bladder and rectum obtained from the two segmentation techniques (P >0.05). The correlation between the slice volumes of all organs of interest provided by the region growing technique and manual tracing was very good to excellent (prostate, r2=0.84; bladder, r2=0.93; rectum, r2=0.85). The intra- and interobserver variation for prostate, bladder and rectum volume segmentation was found to be lower with the region growing technique than with manual tracing. c) Phantom experiments revealed that the accuracy of the random marking technique and manual planimetry in measuring volume on MR imaging data was approximately the same. In vivo splenic volume measurements derived from both volumetric techniques were highly correlated (r2=0.99, P<0.0001). The random marking technique was more reproducible than the manual planimetry in terms of intra- and interobserver variation. Moreover, the proposed volumetric technique was almost twice as fast as the manual planimetry. Conclusions : Three - dimensional images of paranasal sinuses obtained from the volume rendering technique may be used for the evaluation of bony abnormalities and preoperative planning. The region growing tecnique for image segmentation during treatment planning of prostate cancer and the random marking technique for estimation of splenic volume may provide advantages over the conventional techniques used in current practice.
|
Εφαρμογή σύγχρονων μεθόδων ανασύνθεσης και επεξεργασίας εικόνων στην υπολογιστική και μαγνητική τομογραφία
