Automatic Extraction of the Tooth Outline on the Intra-oral Radiograph Using Wavelet Transforms

DOI 21 References
  • SHIMIZU Toshitaka
    Department of Oral and Maxillofacial Radiology, Faculty of Dental Science, Kyushu University Medical corporation Midorigaoka Dental Clinic
  • TOKUMORI Kenji
    Department of Oral and Maxillofacial Radiology, Faculty of Dental Science, Kyushu University
  • YOSHIURA Kazunori
    Department of Oral and Maxillofacial Radiology, Faculty of Dental Science, Kyushu University
  • KANDA Shigenobu
    Department of Oral and Maxillofacial Radiology, Faculty of Dental Science, Kyushu University

Bibliographic Information

Other Title
  • Ｗａｖｅｌｅｔ変換を利用した歯のＸ線画像の外形線自動抽出に関する研究

Search this article

Abstract

Purpose: This study developed a new method of automatically extracting the tooth outline on digital intra-oral radiographs using Wavelet transformation.<br>Materials and methods: To confirm the precision of this method, a radiograph of a ball phantom was obtained with a DenOptics system. The tube voltage was fixed, and exposure time was changed. Outlines of the phantom were extracted by NIH-Image J with or without Wavelet transformation, or by manual-tracing by the observers. These three types of outline images were compared with the actual size of the phantom.<br>To confirm the reliability of this method, radiographs of the 27 natural extracted teeth and Mix-DP phantom were obtained with three digital intra-oral radiography systems (Compuray, Dixel, and Sens-A-Ray). Outlines of the tooth contours were extracted with or without Wavelet transformation, or by manual-tracing by the observers. These three types of outline images were compared and evaluated.<br>Result: The outline of a ball phantom extracted with Wavelet transformation was superior to those extracted without Wavelet transformation or by manual-tracing by the observers, and the outline was very close to the actual size. For tooth outline extraction, this method was superior to the conventional method and close to that by manual-tracing by the observers. There were no differences in the outlines extracted by this method among the three digital systems.<br>Conclusion: Outline extraction using Wavelet transformation was useful for automatically extracting tooth outlines. This method may be applicable to computed assisted diagnosis system for proximal caries.

Journal

  • Shika Hoshasen

    Shika Hoshasen 44 (3), 161-168, 2004

    Japanese Society for Oral and Maxillofacial Radiology

References(21)*help

See more

Keywords

Details 詳細情報について

Report a problem

Back to top