Fully automatic segmentation of paraspinal muscles from 3D torso CT images via multi-scale iterative random forest classifications
- 214 Downloads
To develop and validate a fully automatic method for segmentation of paraspinal muscles from 3D torso CT images.
We propose a novel learning-based method to address this challenging problem. Multi-scale iterative random forest classifications with multi-source information are employed in this study to speed up the segmentation and to improve the accuracy. Here, multi-source images include the original torso CT images and later also the iteratively estimated and refined probability maps of the paraspinal muscles. We validated our method on 20 torso CT data with associated manual segmentation. We randomly partitioned the 20 CT data into two evenly distributed groups and took one group as the training data and the other group as the test data.
The proposed method achieved a mean Dice coefficient of 93.0%. It took on average 46.5 s to segment a 3D torso CT image with the size ranging from \(512 \times 512 \times 802\) voxels to \(512 \times 512 \times 1031\) voxels.
Our fully automatic, learning-based method can accurately segment paraspinal muscles from 3D torso CT images. It generates segmentation results that are better than those achieved by the state-of-the-art methods.
KeywordsParaspinal muscles CT Segmentation Random forest
This work was supported in part by a JSPS Grant-in-Aid for Scientific Research on Innovative Areas (Multidisciplinary Computational Anatomy, #26108005 and #17H05301), JAPAN.
Compliance with ethical standards
Conflict of interest
The authors declare that they have no conflict of interest.
Informed consent was obtained from all individuals included in the study.
- 4.Dubuisson M, Jain A (1994) A modified hausdorff distance for object matching. In: Proceedings of international conference on pattern recognition (ICPR). pp 566–568Google Scholar
- 9.Kamiya N, Zhou X, Chen H, Hara T, Hoshi H, Yokoyama R, Kanematsu M, Fujita H (2009) Automated recognition of the psoas major muscles on X-ray CT images. In: Proceedings of IEEE-EMBC 2009. pp 3557–3560Google Scholar
- 10.Kamiya N, Zhou X, Chen H, Muramatsu C, Hara T, Yokoyama R, Kanematsu M, Hoshi H, Fujita H (2011) Automated segmentation of recuts abdominis muscle using shape model in X-ray CT images. In: Proceedings of IEEE-EMBC 2011. pp 7993–7996Google Scholar
- 12.Kume M, Kamiya N, Zhou X, Kato H, Chen H, Muramatsu C, Hara T, Miyoshi T, Matsuo M, Fujita H (2017) Automated recognition of the erector spinae muscle based on deep CNN at the level of the twelfth thoracic vertebrae in torso CT images. In: Proceedings of the 36th JAMIT annual meetingGoogle Scholar
- 14.Long J, Shelhamer E, Darrell T (2015) Fully convolutional networks for semantic segmentation. In: Proceedings of 2015 IEEE conference on computer vision and pattern recognition (CVPR 2015). pp 3431–3440Google Scholar
- 16.Nimura Y, Deguchi D, Kitasaka T, Mori K, Suenaga Y (2008) Pluto: a common platform for computer-aided diagnosis. Med Imaging Technol 26(3):187–191Google Scholar
- 17.Ogier A, Sdika M, Foure A, Le Troter A, Bendahan D (2017) Individual muscle segmentation in MR images: A 3D propagation through 2D non-linear registration approaches. In: Proceedings of IEEE-EMBC 2017. pp 317–320Google Scholar
- 24.Simonyan K, Zisserman A (2014) Very deep convolutional networks for large-scale image recognition. arXiv:1409.1556
- 26.Viola P, Jones M (2001) Rapid object detection using a boosted cascade of simple features. In: Proceedings of 2001 CVPR conference. IEEE pp 511–518Google Scholar
- 29.Wei Y, Xu B, Tao X, Qu J (2015) Paraspinal muscle segmentation in CT images using a single atlas. In: Proceedings of IEEE international conference on progress in informatics and computing (IPC). pp 211–215Google Scholar