Presentation 2023-06-15
A New Detection Method of Calcification Regions in Dental Panoramic Radiograph and Application of MTL in Its Learning
Taito Murano, Mitsuji Muneyasu, Soh Yoshida, Akira Asano, Keiichi Uchida,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Calcification regions, considered a sign of atherosclerosis, are sometimes observed in the carotid arteries in dental panoramic radiographs. For this purpose, a semantic segmentation-based method has been proposed to detect calcification regions in dental panoramic radiographs automatically. Since the output of semantic segmentation is an image with a label probability, it should determine the presence of calcification regions from the image. This paper proposes a new detector that combines a segmentation model (TransCaraNet) and a convolutional neural network (Res2Net). We also propose the application of Multi-Task Learning (MTL) to this model. The MTL helps the model efficiently learn two tasks with different characteristics, i.e., the segmentation and discriminant tasks. The proposed method can achieve highly accurate detection. As an experiment, we apply the proposed method to an actual dental panoramic radiograph and show the effectiveness of the proposed method from the results.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) calcification region / automatic detection / vascular disease / semantic segmentation / convolutional neural network / multi-task learning
Paper # SIS2023-2
Date of Issue 2023-06-08 (SIS)

Conference Information
Committee SIS / IPSJ-AVM
Conference Date 2023/6/15(2days)
Place (in Japanese) (See Japanese page)
Place (in English) NIT, Matsue College
Topics (in Japanese) (See Japanese page)
Topics (in English) Intelligent Multimedia Systems, Applied Embedded Systems, Three-Dimensional Image Technology (3DIT), etc.
Chair Tomoaki Kimura(Kanagawa Inst. of Tech.) / Sei Naito(KDDI Research, Inc.)
Vice Chair Naoto Sasaoka(Tottori Univ.) / Hakaru Tamukoh(Kyushu Inst. of Tech.)
Secretary Naoto Sasaoka(Kanagawa Inst. of Tech.) / Hakaru Tamukoh(Kansai Univ.) / (NTT)
Assistant Yuichiro Tanaka(kyushu Inst. of Tech.) / Yosuke Sugiura(Saitama Univ.)

Paper Information
Registration To Technical Committee on Smart Info-Media Systems / Special Interest Group on Audio Visual and Multimedia Information Processing
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) A New Detection Method of Calcification Regions in Dental Panoramic Radiograph and Application of MTL in Its Learning
Sub Title (in English)
Keyword(1) calcification region
Keyword(2) automatic detection
Keyword(3) vascular disease
Keyword(4) semantic segmentation
Keyword(5) convolutional neural network
Keyword(6) multi-task learning
1st Author's Name Taito Murano
1st Author's Affiliation Kansai University(Kansai Univ.)
2nd Author's Name Mitsuji Muneyasu
2nd Author's Affiliation Kansai University(Kansai Univ.)
3rd Author's Name Soh Yoshida
3rd Author's Affiliation Kansai University(Kansai Univ.)
4th Author's Name Akira Asano
4th Author's Affiliation Kansai University(Kansai Univ.)
5th Author's Name Keiichi Uchida
5th Author's Affiliation Matsumoto Dental University Hospital,(Matsumoto Dental Univ. Hospital,)
Date 2023-06-15
Paper # SIS2023-2
Volume (vol) vol.123
Number (no) SIS-77
Page pp.pp.7-12(SIS),
#Pages 6
Date of Issue 2023-06-08 (SIS)