Multiple training of neural networks for automatic spine segmentation

doi:10.26565/2304-6201-2024-62-04

Vladyslav Koniukhov Anatolii Pidhornyi Institute of Mechanical Engineering Problems of NAS of Ukraine, 2/10, Pozharskyi str., Kharkiv, 61046, Ukraine https://orcid.org/0009-0007-0256-1388
Oleg Morgun "Laboratory of X-ray Medical Equipment" LTD, Dostoevsky str. 1, Kharkiv, 61102, Ukraine https://orcid.org/0009-0005-6157-9110
Kostyantyn Nemchenko V. N. Karazin Kharkiv National University Maydan Svobody, 4, Kharkiv, 61022, Ukraine https://orcid.org/0000-0002-0734-942X

DOI: https://doi.org/10.26565/2304-6201-2024-62-04

Keywords: artificial intelligence, machine learning, image recognition, neural network, ensemble of neural networks, morphometry

Abstract

Actuality
Preventive and diagnostic studies of the presence of bone diseases require morphometric studies of X-ray images of the chest area. Nowadays, artificial intelligence methods are increasingly being used to solve such problems. The main difficulties of this task are related to the fact that X-ray images have quality limitations, for example, in terms of signal-to-noise ratio or contrast. For this reason, the application of standard methods of image recognition or automatic diagnosis becomes impossible. These difficulties have led to the fact that there is currently a fairly large number of works in this field, but the results of most of them are insufficient for practical use.

Goal
Investigate the possibility of using artificial intelligence in the segmentation of medical images for the purpose of automatic diagnosis of diseases of the human bone system.

Research methods
The ensemble method of X-ray image segmentation has been used in the study. The baseline of training data was created on the basis of X-ray images taken from open sources. The total number of images is 183. The initial data was modified according to the requirements necessary for model training. All images were converted to grayscale and resized to 256x256 pixels.

Results
Using this method in the two test cases resulted in an improvement in accuracy from 0.543 to 0.820 for the first snapshot and from 0.725 to 0.923 for the second snapshot.

Conclusions
We have proposed and investigated the application of the methodology of using an ensemble of reusable neural networks for automatic segmentation of a certain area of the spine, namely the Th8-Th11 spine region. The application of this method allowed obtaining more stable and accurate predictions for the desired spine regions, even for images with high noise levels.

Downloads

Download data is not yet available.

Author Biographies

Vladyslav Koniukhov, Anatolii Pidhornyi Institute of Mechanical Engineering Problems of NAS of Ukraine, 2/10, Pozharskyi str., Kharkiv, 61046, Ukraine

PhD student

Oleg Morgun, "Laboratory of X-ray Medical Equipment" LTD, Dostoevsky str. 1, Kharkiv, 61102, Ukraine

Ph.D., director

Kostyantyn Nemchenko, V. N. Karazin Kharkiv National University Maydan Svobody, 4, Kharkiv, 61022, Ukraine

D. of Sc., head of the department

References

/

References

G. Guglielmi, D. Diacinti, C. van Kuijk, F. Aparisi, C. Krestan, J. E. Adams, and T. M. Link, "Vertebral morphometric: current methods and recent advances," European Radiology, p. 14, 2008. https://link.springer.com/article/10.1007/s00330-008-0899-8?error=cookies_not_supported&code=d445f129-d830-4a13-9590-5989bcf5141c

S. Li and J. Yao, Spinal Imaging and Image Analysis, vol. 18, p. 507, 2015. https://link.springer.com/book/10.1007/978-3-319-12508-4

S. Ebrahimi, L. Gajny, W. Skalli, and E. Angelini, "Vertebral corners detection on sagittal X-rays based on shape modelling, random forest classifiers and dedicated visual features," Computer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization, vol. 7, no. 2, pp. 132-144, 2018. https://hal.science/hal-02181802/file/IBHGC_CMBBEIV_2018_Ebrahimi.pdf

L. R. Long and G. R. Thoma, "Segmentation and feature extraction of cervical spine x-ray images," in Proc. SPIE 3661, Medical Imaging 1999: Image Processing, May 1999. https://ui.adsabs.harvard.edu/abs/1999SPIE.3661.1037L/abstract

O. Ronneberger, P. Fischer, and T. Brox, "U-Net: Convolutional Networks for Biomedical Image Segmentation," 2015. https://arxiv.org/abs/1505.04597

S. M. M. R. Al Arif, "Fully automatic image analysis framework for cervical vertebra in X-ray images," Doctoral thesis, City, University of London, 2018. https://pubmed.ncbi.nlm.nih.gov/29477438/

K. C. Kim, H. C. Cho, T. J. Jang, J. M. Choi, and J. K. Seo, "Automatic detection and segmentation of lumbar vertebrae from X-ray images for compression fracture evaluation," arXiv:1904.07624v1 [physics.med-ph], Apr. 2019. https://arxiv.org/abs/1904.07624

Y. Chen, Y. Mo, A. Readie, G. Ligozio, T. Coroller, and B. W. Papie, "VertXNet: Automatic Segmentation and Identification of Lumbar and Cervical Vertebrae from Spinal X-ray Images," arXiv:2207.05476v1 [eess.IV], Jul. 2022. https://arxiv.org/abs/2207.05476

L. R. Dice, "Measures of the amount of ecologic association between species," Ecology, p. 297-302, 1945. https://esajournals.onlinelibrary.wiley.com/doi/10.2307/1932409

"Vindr.ai Datasets: SpineXR." [Online]. Available: https://vindr.ai/datasets/spinexr. [Accessed: 16-Jun-2024].

G. Guglielmi, D. Diacinti, C. van Kuijk, F. Aparisi, C. Krestan, J.E. Adams, T.M. Link. Vertebral morphometric: current methods and recent advances. European Radiology. – 2008. – P. 14. https://link.springer.com/article/10.1007/s00330-008-0899-8?error=cookies_not_supported&code=d445f129-d830-4a13-9590-5989bcf5141c

Shuo Li, Jianhua Yao. Spinal Imaging and Image Analysis. – 2015. – V. 18. – P. 507. https://link.springer.com/book/10.1007/978-3-319-12508-4

Shahin Ebrahimi, Laurent Gajny, Wafa Skalli, Elsa Angelini. Vertebral corners detection on sagittal X-rays based on shape modelling, random forest classifiers and dedicated visual features. Computer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization, 2018, 7 (2), pp. 132-144. https://hal.science/hal-02181802/file/IBHGC_CMBBEIV_2018_Ebrahimi.pdf

L. Rodney Long, George R. Thoma, "Segmentation and feature extraction of cervical spine x-ray images," Proc. SPIE 3661, Medical Imaging 1999: Image Processing, (21 May 1999). https://ui.adsabs.harvard.edu/abs/1999SPIE.3661.1037L/abstract

Olaf Ronneberger, Philipp Fischer, Thomas Brox. U-Net: Convolutional Networks for Biomedical Image Segmentation. – 2015. – P. 8. https://arxiv.org/abs/1505.04597

Al Arif, S.M.M.R. (2018). Fully automatic image analysis framework for cervical vertebra in X-ray images. (Unpublished Doctoral thesis, City, University of London). https://pubmed.ncbi.nlm.nih.gov/29477438/

Kang Cheol Kim, Hyun Cheol Cho, Tae Jun Jang, Jong Mun Choi, Jin Keun Seo. Automatic detection and segmentation of lumbar vertebrae from X-ray images for compression fracture evaluation. arXiv:1904.07624v1 [physics.med-ph] 16 Apr 2019. https://arxiv.org/abs/1904.07624

Yao Chen, Yuanhan Mo, Aimee Readie, Gregory Ligozio, Thibaud Coroller, Bartolomiej W. Papie, VertXNet: Automatic Segmentation and Identification of Lumbar and Cervical Vertebrae from Spinal X-ray Images, arXiv:2207.05476v1 [eess.IV] 12 Jul 2022. https://arxiv.org/abs/2207.05476

Dice, Lee.R. Measures of the amount of ecologic association between species. Ecology. – 1945. – P. 297-302. https://esajournals.onlinelibrary.wiley.com/doi/10.2307/1932409

"Vindr.ai Datasets: SpineXR." [Online]. URL: https://vindr.ai/datasets/spinexr (дата звернення: 16.24.2024)