Abstract
Computed tomography (CT) scans are an excellent way of capturing the details of the abdominal muscles. Several abnormalities including hernia, tumour, and neuro-muscular diseases can be identified and estimated by CT scans, including the representation of muscle loss. Radiologists carefully segment and review every CT slice to monitor these abnormalities. This is a meticulous, painstaking, and time-consuming process. This is especially true for the abdomen, which has a lot of variations and complexities in the wall and surrounding tissues. To solve this, we annotated 451 image-mask pairs from DICOM files of the abdominal muscles of size \(512 \times 512\) pixels. In this work, the trade-off between computation cost and image loss is estimated by splitting the imageries into two categories of size: \(128 \times 128\) pixels and \(512 \times 512\) pixels. To sustain the original image size, the authors create a resized version of the base U-Net and train it from scratch. Furthermore, two dataset categories are created to show the data-intensive quality of the models (One double in size compared to the other). To train the system using these datasets, two architectures U-Net and resized U-Net are used which are then compared through the introduction of a new loss function—“complete loss”. The resized U-Net performs better than the standard U-Net by displaying a mean complete loss of 3.52 and 10.79% for the validation and test sets, respectively. The data dependency of models is clearly shown as well.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Ronneberger O, Fischer P, Brox T (2015) U-net: convolutional networks for biomedical image segmentation. In: International conference on medical image computing and computer-assisted intervention. Springer, Cham, pp 234–241
Boita J, van Engen RE, Mackenzie A, Tingberg A, Bosmans H, Bolejko A, Zackrisson S, Wallis MG, Ikeda DM, Van Ongeval C, Pijnappel R, VISUAL Group Jansen F, Duijm L, de Bruin H, Andersson I, Behmer C, Taylor K, Kilburn-Toppin F, van Goethem M, Prevos R, Salem N, Pal S (2021) How does image quality affect radiologists’ perceived ability for image interpretation and lesion detection in digital mammography? Eur Radiol 31:5335–5343
Gonzalez RC, Woods RE, Eddins SL (2004) Digital image processing using MATLAB. Pearson Education India
Pal NR, Pal SK (1993) A review on image segmentation techniques. Pattern Recogn 26(9):1277–1294
LeCun Y, Bengio Y, Hinton G (2015) Deep learning. Nature 521(7553):436–444
Shin HC, Roth HR, Gao M, Lu L, Xu Z, Nogues I, Summers RM (2016) Deep convolutional neural networks for computer-aided detection: CNN architectures, dataset characteristics, and transfer learning. IEEE Trans Med Imaging 35(5):1285–1298
Krizhevsky A, Sutskever I, Hinton GE (2017) Imagenet classification with deep convolutional neural networks. Commun ACM 60(6):84–90
Long J, Shelhamer E, Darrell T (2015) Fully convolutional networks for semantic segmentation. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 3431–3440
Zhou Y, Xie L, Shen W, Wang Y, Fishman EK, Yuille AL (2017) A fixed-point model for pancreas segmentation in abdominal CT scans. In: International conference on medical image computing and computer-assisted intervention. Springer, Cham, pp 693–701
Roth HR, Oda H, Hayashi Y, Oda M, Shimizu N, Fujiwara M, Misawa K, Mori K (2017) Hierarchical 3D fully convolutional networks for multi-organ segmentation. arXiv:1704.06382
Chlebus G, Schenk A, Moltz JH, van Ginneken B, Hahn HK, Meine H (2018) Automatic liver tumor segmentation in CT with fully convolutional neural networks and object-based postprocessing. Sci Rep 8(1):1–7
Jaeger PF, Kohl SA, Bickelhaupt S, Isensee F, Kuder TA, Schlemmer HP, Maier-Hein KH (2020) Retina U-net: embarrassingly simple exploitation of segmentation supervision for medical object detection. In: Machine learning for health workshop. PMLR, pp 171–183
Yushkevich PA, Piven J, Hazlett HC, Smith RG, Ho S, Gee JC, Gerig G (2006) User-guided 3D active contour segmentation of anatomical structures: significantly improved efficiency and reliability. Neuroimage 31(3):1116–1128
Kirkland EJ, Kirkland EJ (2010) Bilinear interpolation. In: Advanced computing in electron microscopy, pp 261–263
Roth HR, Shen C, Oda H, Oda M, Hayashi Y, Misawa K, Mori K (2018) Deep learning and its application to medical image segmentation. Med Imaging Technol 36(2):63–71
Rister B, Yi D, Shivakumar K, Nobashi T, Rubin DL (2020) CT-ORG, a new dataset for multiple organ segmentation in computed tomography. Sci Data 7(1):1–9
Kingma DP, Ba J (2014) Adam: a method for stochastic optimization. arXiv:1412.6980
Sharma N, Aggarwal LM (2010) Automated medical image segmentation techniques. J Med Phys 35(1):3
Acknowledgements
This paper and the research would not have been possible without the help of my exceptional mentor Maniappan R and Curium Life for providing us with the datasets. His valuable suggestions during the planning and development of this research work helped us progress as scheduled.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
About this paper
Cite this paper
Bisleri, A.P., Kumar, A.S., Bhargav, A.A., Varun, S., Kasturi, N. (2023). Size Matters: Exploring the Impact of Model Architecture and Dataset Size on Semantic Segmentation of Abdominal Wall Muscles in CT Scans. In: Ranganathan, G., Papakostas, G.A., Rocha, Á. (eds) Inventive Communication and Computational Technologies. ICICCT 2023. Lecture Notes in Networks and Systems, vol 757. Springer, Singapore. https://doi.org/10.1007/978-981-99-5166-6_42
Download citation
DOI: https://doi.org/10.1007/978-981-99-5166-6_42
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-99-5165-9
Online ISBN: 978-981-99-5166-6
eBook Packages: Intelligent Technologies and RoboticsIntelligent Technologies and Robotics (R0)