The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. ex. Some numerals are expressed as "XNUMX".
Copyrights notice
The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. Copyrights notice
Analisis cephalometry kuantitatif yang tepat dan automatik adalah sangat penting dalam ortodontik. Langkah asas untuk analisis cephalometry adalah untuk menganotasi tanda tempat anatomi yang berminat pada imej X-ray. Kaedah automatik berbantukan komputer masih menjadi topik terbuka pada masa kini. Dalam makalah ini, kami mencadangkan pendekatan kasar-ke-halus berasaskan pembelajaran mendalam yang cekap untuk merealisasikan pengesanan mercu tanda yang tepat. Dalam langkah pengesanan kasar, kami melatih model transformasi boleh ubah bentuk berasaskan pembelajaran mendalam dengan menggunakan sampel latihan. Kami mendaftarkan imej ujian kepada imej rujukan (satu imej latihan) menggunakan model terlatih untuk meramalkan lokasi mercu tanda kasar pada imej ujian. Oleh itu, kawasan menarik (ROI) yang termasuk tanda tempat boleh ditemui. Dalam langkah pengesanan halus, kami menggunakan rangkaian neural convolutional deep terlatih (CNN), untuk mengesan tanda tempat dalam tampung ROI. Untuk setiap tanda tempat, terdapat satu rangkaian saraf yang sepadan, yang secara langsung melakukan regresi ke koordinat mercu tanda. Langkah halus boleh dianggap sebagai langkah pemurnian atau penalaan halus berdasarkan langkah pengesanan kasar. Kami mengesahkan kaedah yang dicadangkan pada set data awam daripada cabaran besar Simposium Antarabangsa mengenai Pengimejan Bioperubatan (ISBI) 2015. Berbanding dengan kaedah tercanggih, kami bukan sahaja mencapai ketepatan pengesanan setanding (min ralat jejarian adalah kira-kira 1.0-1.6mm), tetapi juga memendekkan sebahagian besar masa pengiraan (4 saat setiap imej).
Yu SONG
Ritsumeikan University
Xu QIAO
Shandong University
Yutaro IWAMOTO
Ritsumeikan University
Yen-Wei CHEN
Ritsumeikan University
Yili CHEN
Zhejiang University
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Salinan
Yu SONG, Xu QIAO, Yutaro IWAMOTO, Yen-Wei CHEN, Yili CHEN, "An Efficient Deep Learning Based Coarse-to-Fine Cephalometric Landmark Detection Method" in IEICE TRANSACTIONS on Information,
vol. E104-D, no. 8, pp. 1359-1366, August 2021, doi: 10.1587/transinf.2021EDP7001.
Abstract: Accurate and automatic quantitative cephalometry analysis is of great importance in orthodontics. The fundamental step for cephalometry analysis is to annotate anatomic-interested landmarks on X-ray images. Computer-aided automatic method remains to be an open topic nowadays. In this paper, we propose an efficient deep learning-based coarse-to-fine approach to realize accurate landmark detection. In the coarse detection step, we train a deep learning-based deformable transformation model by using training samples. We register test images to the reference image (one training image) using the trained model to predict coarse landmarks' locations on test images. Thus, regions of interest (ROIs) which include landmarks can be located. In the fine detection step, we utilize trained deep convolutional neural networks (CNNs), to detect landmarks in ROI patches. For each landmark, there is one corresponding neural network, which directly does regression to the landmark's coordinates. The fine step can be considered as a refinement or fine-tuning step based on the coarse detection step. We validated the proposed method on public dataset from 2015 International Symposium on Biomedical Imaging (ISBI) grand challenge. Compared with the state-of-the-art method, we not only achieved the comparable detection accuracy (the mean radial error is about 1.0-1.6mm), but also largely shortened the computation time (4 seconds per image).
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2021EDP7001/_p
Salinan
@ARTICLE{e104-d_8_1359,
author={Yu SONG, Xu QIAO, Yutaro IWAMOTO, Yen-Wei CHEN, Yili CHEN, },
journal={IEICE TRANSACTIONS on Information},
title={An Efficient Deep Learning Based Coarse-to-Fine Cephalometric Landmark Detection Method},
year={2021},
volume={E104-D},
number={8},
pages={1359-1366},
abstract={Accurate and automatic quantitative cephalometry analysis is of great importance in orthodontics. The fundamental step for cephalometry analysis is to annotate anatomic-interested landmarks on X-ray images. Computer-aided automatic method remains to be an open topic nowadays. In this paper, we propose an efficient deep learning-based coarse-to-fine approach to realize accurate landmark detection. In the coarse detection step, we train a deep learning-based deformable transformation model by using training samples. We register test images to the reference image (one training image) using the trained model to predict coarse landmarks' locations on test images. Thus, regions of interest (ROIs) which include landmarks can be located. In the fine detection step, we utilize trained deep convolutional neural networks (CNNs), to detect landmarks in ROI patches. For each landmark, there is one corresponding neural network, which directly does regression to the landmark's coordinates. The fine step can be considered as a refinement or fine-tuning step based on the coarse detection step. We validated the proposed method on public dataset from 2015 International Symposium on Biomedical Imaging (ISBI) grand challenge. Compared with the state-of-the-art method, we not only achieved the comparable detection accuracy (the mean radial error is about 1.0-1.6mm), but also largely shortened the computation time (4 seconds per image).},
keywords={},
doi={10.1587/transinf.2021EDP7001},
ISSN={1745-1361},
month={August},}
Salinan
TY - JOUR
TI - An Efficient Deep Learning Based Coarse-to-Fine Cephalometric Landmark Detection Method
T2 - IEICE TRANSACTIONS on Information
SP - 1359
EP - 1366
AU - Yu SONG
AU - Xu QIAO
AU - Yutaro IWAMOTO
AU - Yen-Wei CHEN
AU - Yili CHEN
PY - 2021
DO - 10.1587/transinf.2021EDP7001
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E104-D
IS - 8
JA - IEICE TRANSACTIONS on Information
Y1 - August 2021
AB - Accurate and automatic quantitative cephalometry analysis is of great importance in orthodontics. The fundamental step for cephalometry analysis is to annotate anatomic-interested landmarks on X-ray images. Computer-aided automatic method remains to be an open topic nowadays. In this paper, we propose an efficient deep learning-based coarse-to-fine approach to realize accurate landmark detection. In the coarse detection step, we train a deep learning-based deformable transformation model by using training samples. We register test images to the reference image (one training image) using the trained model to predict coarse landmarks' locations on test images. Thus, regions of interest (ROIs) which include landmarks can be located. In the fine detection step, we utilize trained deep convolutional neural networks (CNNs), to detect landmarks in ROI patches. For each landmark, there is one corresponding neural network, which directly does regression to the landmark's coordinates. The fine step can be considered as a refinement or fine-tuning step based on the coarse detection step. We validated the proposed method on public dataset from 2015 International Symposium on Biomedical Imaging (ISBI) grand challenge. Compared with the state-of-the-art method, we not only achieved the comparable detection accuracy (the mean radial error is about 1.0-1.6mm), but also largely shortened the computation time (4 seconds per image).
ER -