Journal of Korea Multimedia Society (한국멀티미디어학회논문지)

Korea Multimedia Society (한국멀티미디어학회)
권호 표지
DOI QR코드

DOI QR Code

Automated Ulna and Radius Segmentation model based on Deep Learning on DEXA

DEXA에서 딥러닝 기반의 척골 및 요골 자동 분할 모델

  • Kim, Young Jae (Dept. of Biomedical Engineering, Gachon University College of Medicine) ;
  • Park, Sung Jin (Dept. of Biomedical Engineering, Gachon University College of Medicine) ;
  • Kim, Kyung Rae (Dept. of Biomedical Engineering, Gachon University College of HealthScience) ;
  • Kim, Kwang Gi (Dept. of Biomedical Engineering, Gachon University College of Medicine, Dept. of Biomedical Engineering, Gachon University College of HealthScience)
  • Received : 2018.08.20
  • Accepted : 2018.11.13
  • Published : 2018.12.31
Download PDF
⟨ Previous Next ⟩

Abstract

The purpose of this study was to train a model for the ulna and radius bone segmentation based on Convolutional Neural Networks and to verify the segmentation model. The data consisted of 840 training data, 210 tuning data, and 200 verification data. The learning model for the ulna and radius bone bwas based on U-Net (19 convolutional and 8 maximum pooling) and trained with 8 batch sizes, 0.0001 learning rate, and 200 epochs. As a result, the average sensitivity of the training data was 0.998, the specificity was 0.972, the accuracy was 0.979, and the Dice's similarity coefficient was 0.968. In the validation data, the average sensitivity was 0.961, specificity was 0.978, accuracy was 0.972, and Dice's similarity coefficient was 0.961. The performance of deep convolutional neural network based models for the segmentation was good for ulna and radius bone.

Keywords

MTMDCW_2018_v21n12_1407_f0001.png 이미지

Fig. 1. Example of DEXA image (a) high energy image, (b) low energy image.

MTMDCW_2018_v21n12_1407_f0002.png 이미지

Fig. 2. Pre-processing process in low energy image. The position of the crop area was calculated using the line profile.

MTMDCW_2018_v21n12_1407_f0003.png 이미지

Fig. 3. U-net architecture consisted with convolutional encoding and decoding units.

MTMDCW_2018_v21n12_1407_f0004.png 이미지

Fig. 4. Result of post-processing algorithm (a) original image, (b) False positive image, (c) Result image with post-processing.

MTMDCW_2018_v21n12_1407_f0005.png 이미지

Fig. 5. Comparison of the segmentation results between the deep learning and manual. (a)original images, (b)manual segmentation results, (c)deep learning segmentation results.

MTMDCW_2018_v21n12_1407_f0006.png 이미지

Fig. 6. Scatter plots comparing the manual and the DL area measurements. (a) Manual and Deep Learning in Train data, (b) Manual and Deep Learning in Test data.

MTMDCW_2018_v21n12_1407_f0007.png 이미지

Fig. 7. Bland-Altman plots comparing the manual and the DL area measurements. (a) Manual and Deep Learning in Train data, (b) Manual and Deep Learning in Test data.

MTMDCW_2018_v21n12_1407_f0008.png 이미지

Fig. 8. Comparison of results according to learning rate (a)original image, (b)result by learning rate 0.001, (c)result by learning rate 0.0001, (d)result by learning rate 0.00001

Table 1. Conditional probability results of the trained segmentation model.

MTMDCW_2018_v21n12_1407_t0001.png 이미지

Table 2. Verification and comparison of the deep learning and manual area measurements

MTMDCW_2018_v21n12_1407_t0002.png 이미지

References

  1. J.A. Kanis, “Assessment of Fracture Risk and Its Application to Screening for Postmenopausal Osteoporosis: Synopsis of a WHO Report,” Osteoporosis International, Vol. 4, No. 6, pp. 368-381, 1994. https://doi.org/10.1007/BF01622200
  2. A. Klibanski, L. Adams-campbell, T. Bassford, S.N. Blair, S.D. Boden, K. Dickersin et al., “Osteoporosis Prevention, Diagnosis, and Therapy,” Journal of the American Medical Association, Vol. 285, No. 6, pp. 785-95, 2001. https://doi.org/10.1001/jama.285.6.785
  3. E. Hernlund, A. Svedbom, M. Ivergard, J. Compston, C. Cooper, J. Stenmark, et al., “Osteoporosis in the European Union: Medical Management, Epidemiology and Economic Burden,” Archives of Osteoporos, Vol. 8, No. 1-2, pp. 136, 2013. https://doi.org/10.1007/s11657-013-0136-1
  4. J.W. Kwon, S.I. Cho, Y.B. Ahn, and Y.M. Ro, “Bone Region Extraction by Dual Energy X-ray Absorbtion Image Decomposition,” Journal of Korea Multimed Society, Vol. 12, No. 9, pp. 1233-1241, 2009.
  5. N.B. Watts, E.M. Lewiecki, P.D. Miller, and S. Baim, “National Osteoporosis Foundation 2008 Clinician's Guide to Prevention and Treatment of Osteoporosis and the World Health Organization Fracture Risk Assessment Tool (FRAX): What They Mean to the Bone Densitometrist and Bone Technologist,” Journal of Clinical Densitometry, Vol. 11, No. 4, pp. 473-477, 2008. https://doi.org/10.1016/j.jocd.2008.04.003
  6. R. Michael and M. Clung, “Prevention and Management of Osteoporosis,” Best Practice and Research: Clinical Endocrinology and Metabolism, Vol. 17, No. 1, pp. 53-71, 2003. https://doi.org/10.1016/S1521-690X(02)00080-5
  7. V. Rebuffel and J. Dinten, "Dual-energy X-ray Imaging?: Benefits and Limits," Insight-non-destructive Testing and Condition Monitoring, Vol. 49, No. 10, pp. 589-594, 2006. https://doi.org/10.1784/insi.2007.49.10.589
  8. C. Njeh and D. Hans, The Evaluation of Osteoporosis?: Dual Energy X-ray Absorptiometry and Ultrasound in Clinical Practice, Martin Dunitz Publishers, London, 1999.
  9. A. Krizhevsky, I. Sutskever, and G.E. Hinton, "ImageNet Classification with Deep Convolutional Neural Networks," Proceeding of International Conference on Neural Information Processing Systems, pp. 1-9, 2012.
  10. H.C. Shin, H.R. Roth, M. Gao, L. Lu, Z. Xu, I. Nogues, et al., “Deep Convolutional Neural Networks for Computer-aided Detection: CNN Architectures, Dataset Characteristics and Transfer Learning,” IEEE Transactions on Medical Imaging, Vol. 35, No. 5, pp. 1285-1298, 2017. https://doi.org/10.1109/TMI.2016.2528162
  11. A. Raj, S. Vishwanathan, B. Ajani, K. Krishnan, and H. Agarwal, "Automatic Knee Cartilage Segmentation Using Fully Volumetric Convolutional Neural Networks for Evaluation of Osteoarthritis," Proceeding of IEEE International Symposium on Biomedical Imaging, pp. 851-854, 2018.
  12. M. Havaei, A. Davy, D. Warde-farley, A. Biard, A. Courville, Y. Bengio, et al., "Brain Tumor Segmentation with Deep Neural Networks," Medical Image Analysis, Vol. 35, pp. 18-31, 2017. https://doi.org/10.1016/j.media.2016.05.004
  13. A. Kalinovsky and V. Kovalev, "Lung Image Segmentation Using Deep Learning Methods and Convolutional Neural Networks Lung Image Segmentation Using Deep Learning Methods and Convolutional Neural Networks," Proceeding of International Conference on Pattern Recognition and Information Processing, pp. 21-24, 2016.
  14. B. Norman, V. Pedoia, and S. Majumdar, “Use of 2D U-net Convolutional Neural Networks for Automated Cartilage and Meniscus Segmentation of Knee MR Imaging Data to Determine Relaxometry and Morphometry,” Radiology, Vol. 288, No. 1, pp. 177-185, 2018. https://doi.org/10.1148/radiol.2018172322
  15. O. Cicek, A. Abdulkadir, S.S. Lienkamp, T. Brox, and O. Ronneberger, "3D U-net: Learning Dense Volumetric Segmentation from Sparse Annotation," Proceeding of International Conference on Medical Image Computing and Computer-assisted Intervention, pp. 424-432, 2016.
  16. O. Ronneberger, P. Fischer, and T. Brox, "U-net: Convolutional Networks for Biomedical Image Segmentation," Proceeding of International Conference on Medical Image Computing and Computer-assisted Intervention, pp. 1-8. 2015.
  17. R. Sharma and R. Sharma, “Image Segmentation Using Morphological Operation for Automatic Region Growing,” International Journal of Innovative Research in Computer and Communication Engineering, Vol. 4, No. 6, pp. 5686-5692, 2014.

Cited by

  1. 손목 관절 단순 방사선 영상에서 딥 러닝을 이용한 전후방 및 측면 영상 분류와 요골 영역 분할 vol.41, pp.2, 2018, https://doi.org/10.9718/jber.2020.41.2.94