Takahiro Minamoto

Clinical evaluation of a newly developed method for avoiding artifacts caused by dental fillings on X-ray CT.

To evaluate the clinical usefulness of gantry tilt scanning as an image reconstruction technique for avoiding artifacts caused by metallic dental fillings. Gantry tilt scanning was used with multidetector-row computed tomography for imaging in patients with dental fillings. Using a novel PC-based program, the oblique images obtained were reconstructed to transverse images using nearest neighbor and bilinear interpolation methodologies in order to avoid metallic streak artifacts. Coronal images were reformatted with the reconstructed transverse images, and the continuity of the reconstructed images was evaluated. Gantry tilt scanning was performed in 12 patients with metal artifacts, and the original and reconstructed images were classified into four grades and assessed by two radiologists. Results of the clinical evaluation indicated that the original images with artifacts, only 4% had good image quality in the region around the medial pterygoid muscle, only 8% depicted areas around the internal carotid artery and internal jugular vein, and only 12% could depict the areas around the parotid gland in the clinical evaluation. These values were improved to 60, 96, and 100%, respectively, in the reconstructed transverse images. Gantry tilt scanning as an image reconstruction technique improves image quality and removes most, if not all, artifacts caused by metallic dental fillings. The resulting images can be used in the evaluation of oropharyngeal lesions in patients with dental fillings.

Evaluation of local look diffusion tensor imaging for magnetic resonance tractography of the periprostatic neurovascular bundle

The purpose of this study was to compare diffusion tensor imaging using the local look technique and sensitivity encoding for tractography of the periprostatic neurovascular bundle. We compared the surrounding tissues of the prostate in eight healthy volunteers. The results of tractography in terms of the numbers of fibers and the fractional anisotropy map were evaluated. Distortion was evaluated using the dice similarity coefficient between isotropic diffusion images created from diffusion tensor images and T2-weighted images. The number of lines in tractography was significantly greater in diffusion tensor imaging using the local look technique (p < 0.001). Although there was no difference in image distortion of the prostate between methods, an artifact appeared in the center of the diffusion tensor image using sensitivity encoding. In conclusion, diffusion tensor imaging using the local look technique was superior to that using sensitivity encoding for tractography of the periprostatic neurovascular bundle.