Registration of Two Dental Panoramic Radiographs

In dental panoramic images, the information on physical changes of alveolar bone or jaw bone is very important to diagnose several diseases. To detect such change, it is useful to compare two panoramic x-ray images acquired at different times. These two images are usually acquired with different conditions in terms of the positioning of the dental arch, and thus these images can be impaired from some geometrical changes related to the scale of the panoramic images and deformation of the teeth and jaw bones. As a result of this, it is very hard to make an accurate registration. To cope with this issue, we developed a dedicated image registration method to match these two images by a newly introduced non-rigid transformation method and registration method using the cross-correlation of localized regions. We evaluated our proposed method with several sets of two images acquired with different geometrical conditions. The material evaluated in this study was a skull phantom. The results of these experiments showed the validity and intrinsic ability of our proposed method in clinical examinations.

Removal of the Shadow of Cervical Vertebrae from Panoramic X-Ray Images with a Tomosynthesis Method

The purpose of this study is to remove the shadow of cervical vertebrae from dental panoramic x-ray images with a tomosynthesis method and improve the contrast of details in both the teeth and jaw bones. To measure the shift-amount at each angular position that was required for reconstruction of panoramic x-ray images of the dental arch, strip images of a calibration phantom were acquired. Then, a shift-amount table was prepared from these images, and the other shift-amount table, which was used to reconstruct a panoramic image of the cervical vertebrae, was prepared by inverting the curve of the shift-amount table upside down. Using these two tables, images focused on the dental arch and cervical vertebrae of a patient were made with the original strip data of the patient. The shadow of the cervical vertebrae appearing on the image focused on the dental arch was removed using the two above-mentioned images and blurring functions defined at two focusing geometries. The validity of the proposed method was evaluated with clinically acquired data of two patients. The shadow of the cervical vertebrae was successfully eliminated, and the contrast of the front teeth and detailed structures of the jaw bones was improved. The results of the experiments showed that our proposed method was significantly effective in removing the shadow of the cervical vertebrae from conventional panoramic x-ray images.