Automated Pulmonary Nodule Detection in LDCT Using 3D ResNet and Adaptive Patch Strategy

Dr. Ahmed El-Sayed Mansour; Dr. Sophia Martinez

doi:10.64917/fmcs-001

Authors

Dr. Ahmed El-Sayed Mansour Radiology and Medical Imaging Department, Cairo University, Cairo, Egypt
Dr. Sophia Martinez Division of Pulmonary and Critical Care Medicine, Mayo Clinic, Rochester, MN, USA

DOI:

https://doi.org/10.64917/fmcs-001

Keywords:

Pulmonary Nodule Segmentation, Low-Dose CT, Deep Learning, 3D Convolutional Neural Networks

Abstract

Lung cancer remains a leading cause of cancer-related mortality worldwide [1, 2]. Early detection through low-dose computed tomography (LDCT) screening has been shown to reduce mortality [3, 4, 5]. A critical step in the analysis of LDCT scans for lung cancer screening is the accurate segmentation of pulmonary nodules. Manual segmentation is time-consuming and subject to inter-observer variability. Automated segmentation methods, particularly those leveraging deep learning, offer a promising alternative [13, 14, 15, 16, 17]. This paper proposes a method for automated pulmonary nodule segmentation in LDCT scans utilizing 3D residual networks and a dynamic patch-based sampling strategy. The use of 3D networks is motivated by their ability to capture volumetric context, which is crucial for analyzing 3D medical images [7, 8, 9]. An adaptive patch sampling approach is employed to address the class imbalance inherent in medical image segmentation, where nodules occupy a small fraction of the total volume. We describe the methodology, including data preprocessing using the Lung Image Database Consortium (LIDC-IDRI) dataset [12], the architecture of the 3D residual segmentation network, the dynamic patch sampling strategy, and the training procedure. The potential impact of this approach on improving the accuracy and efficiency of lung cancer screening is discussed.

References

Siegel, R.L.; Miller, K.D.; Jemal, A. Cancer statistics, 2020. CA Cancer J. Clin. 2020, 70, 7–30.

The American Cancer Society Medical and Editorial Content Team. Key Statistics for Lung Cancer. American Cancer Society. Available online: https://www.cancer.org/cancer/types/lung-cancer/about/key-statistics.html (accessed on 21 September 2023).

The American Cancer Society Medical and Editorial Content Team. Lung Cancer Early Detection, Diagnosis, and Staging. The American Cancer Society. Available online: https://www.cancer.org/content/dam/CRC/PDF/Public/8705.00.pdf (accessed on 21 September 2023).

Rampinelli, C.; Origgi, D.; Bellomi, M. Low-dose CT: Technique, reading methods and image interpretation. Cancer Imaging 2013, 12, 548–556.

Tylski, E.; Goyal, M. Low Dose CT for Lung Cancer Screening: The Background, the Guidelines, and a Tailored Approach to Patient Care. Mo. Med. 2019, 116, 414–419.

Ferlay, J.; Ervik, M.; Lam, F.; Colombet, M.; Mery, L.; Piñeros, M. Global Cancer Observatory: Cancer Today. International Agency for Research on Cancer: Lyon, France. Available online: https://gco.iarc.fr/today (accessed on 21 September 2023).

Crespi, L.; Loiacono, D.; Sartori, P. Are 3D better than 2D Convolutional Neural Networks for Medical Imaging Semantic Segmentation? In Proceedings of the 2022 International Joint Conference on Neural Networks (IJCNN), Padua, Italy, 18–23 July 2022; pp. 1–8.

Li, W.; Wang, G.; Fidon, L.; Ourselin, S.; Cardoso, M.J.; Vercauteren, T. On the Compactness, Efficiency, and Representation of 3D Convolutional Networks: Brain Parcellation as a Pretext Task. In Proceedings of the International Conference on Information Processing in Medical Imaging, Boone, NC, USA, 25–30 June 2017.

Vahedifard, F.; Liu, X.; Kocak, M.; Ai, H.A.; Supanich, M.; Marathu, K.K.; Adler, S.; Orouskhani, M.; Byrd, S. Artificial Intelligence in Fetal Resting-State Functional MRI Brain Segmentation: A Comparative Analysis of 3D UNet, VNet, and HighRes-Net Models. arXiv 2023, arXiv:2311.10844.

Logothetis, N.K. What we can do and what we cannot do with fMRI. Nature 2008, 453, 869–878.

Tappeiner, E.; Pröll, S.; Hönig, M.; Raudaschl, P.F.; Zaffino, P.; Spadea, M.F.; Sharp, G.C.; Schubert, R.; Fritscher, K. Multi-organ segmentation of the head and neck area: An efficient hierarchical neural networks approach. Int. J. Comput. Assist. Radiol. Surg. 2019, 14, 745–754.

Armato, S.G.; McLennan, G.; Bidaut, L.; McNitt-Gray, M.F.; Meyer, C.R.; Reeves, A.P.; Zhao, B.; Aberle, D.R.; Henschke, C.I.; Hoffman, E.A.; et al. The Lung Image Database Consortium (LIDC) and Image Database Resource Initiative (IDRI): A completed reference database of lung nodules on CT scans. Med. Phys. 2011, 38, 915–931.

Doi, K. Computer-aided diagnosis in medical imaging: Historical review, current status and future potential. Comput. Med. Imaging Graph. 2007, 31, 198–211.

Jiang, X.; Hu, Z.; Wang, S.; Zhang, Y. Deep Learning for Medical Image-Based Cancer Diagnosis. Cancers 2023, 15, 3608.

Li, J.; Li, P.; Li, H.; Ying, T. Deep Learning-based Semantic Segmentation Methods in Medical Imaging. Highlights Sci. Eng. Technol. 2023, 39, 936–942.

Huang, K. Application of deep learning in medical imaging segmentation. Theor. Nat. Sci. 2023, 17, 27–33.

Marinakis, I.; Karampidis, K.; Papadourakis, G. Pulmonary Nodule Detection, Segmentation and Classification Using Deep Learning: A Comprehensive Literature Review. BioMedInformatics 2024, 4, 2043–2106.

Tang, H.; Zhang, C.; Xie, X. NoduleNet: Decoupled False Positive Reduction for Pulmonary Nodule Detection and Segmentation. In Proceedings of the Medical Image Computing and Computer Assisted Intervention–MICCAI 2019: 22nd International Conference, Shenzhen, China, 13–17 October 2019; pp. 266–274.

Usman, M.; Lee, B.-D.; Byon, S.-S.; Kim, S.-H.; Lee, B.; Shin, Y.-G. Volumetric lung nodule segmentation using adaptive ROI with multi-view residual learning. Sci. Rep. 2020, 10, 12839.

Kido, S.; Kidera, S.; Hirano, Y.; Mabu, S.; Kamiya, T.; Tanaka, N.; Suzuki, Y.; Yanagawa, M.; Tomiyama, N. Segmentation of Lung Nodules on CT Images Using a Nested Three-Dimensional Fully Connected Convolutional Network. Front. Artif. Intell. 2022, 5, 782225.

Qiu, J.; Li, B.; Liao, R.; Mo, H.; Tian, L. A dual-task region-boundary aware neural network for accurate pulmonary nodule segmentation. J. Vis. Commun. Image Represent. 2023, 96, 103909.

Luo, S.; Zhang, J.; Xiao, N.; Qiang, Y.; Li, K.; Zhao, J.; Meng, L.; Song, P. DAS-Net: A lung nodule segmentation method based on adaptive dual-branch attention and shadow mapping. Appl. Intell. 2022, 52, 15617–15631.

Tyagi, S.; Talbar, S.N. CSE-GAN: A 3D conditional generative adversarial network with concurrent squeeze-and-excitation blocks for lung nodule segmentation. Comput. Biol. Med. 2022, 147, 105781.

Ma, X.; Song, H.; Jia, X.; Wang, Z. An improved V-Net lung nodule segmentation model based on pixel threshold separation and attention mechanism. Sci. Rep. 2024, 14, 4743.

Liu, J.; Li, Y.; Li, W.; Li, Z.; Lan, Y. Multiscale lung nodule segmentation based on 3D coordinate attention and edge enhancement. Electron. Res. Arch. 2024, 32, 3016–3037.

Rikhari, H.; Kayal, E.B.; Ganguly, S.; Sasi, A.; Sharma, S.; Antony, A.; Rangarajan, K.; Bakhshi, S.; Kandasamy, D.; Mehndiratta, A. Improving lung nodule segmentation in thoracic CT scans through the ensemble of 3D U-Net models. Int. J. Comput. Assist. Radiol. Surg. 2024, 19, 2089–2099.

Jiang, W.; Zhi, L.; Zhang, S.; Zhou, T. A Dual-Branch Framework With Prior Knowledge for Precise Segmentation of Lung Nodules in Challenging CT Scans. IEEE J. Biomed. Health Inform. 2024, 28, 1540–1551.

Jian, M.; Jin, H.; Zhang, L.; Wei, B.; Yu, H. DBPNDNet: Dual-branch networks using 3DCNN toward pulmonary nodule detection. Med. Biol. Eng. Comput. 2024, 62, 563–573.

Xu, X.; Du, L.; Yin, D. Dual—Branch feature fusion S3D V—Net network for lung nodules segmentation. J. Appl. Clin. Med. Phys. 2024, 25, e14331.

Bbosa, R.; Gui, H.; Luo, F.; Liu, F.; Efio-Akolly, K.; Chen, Y.-P.P. MRUNet-3D: A multi-stride residual 3D UNet for lung nodule segmentation. Methods 2024, 226, 89–101.

Setio, A.A.A.; Traverso, A.; de Bel, T.; Berens, M.S.; Bogaard, C.v.D.; Cerello, P.; Chen, H.; Dou, Q.; Fantacci, M.E.; Geurts, B.; et al. Validation, comparison, and combination of algorithms for automatic detection of pulmonary nodules in computed tomography images: The LUNA16 challenge. Med. Image Anal. 2017, 42, 1–13.

PyLIDC: A Python Interface for the LIDC/IDRI Dataset. Available online: https://github.com/notmatthancock/pylidc (accessed on 11 January 2025).

Wikipedia Contributors. Hounsfield Scale. Available online: https://en.wikipedia.org/w/index.php?title=Hounsfield_scale&oldid=1167604704 (accessed on 18 October 2023).

Pérez-García, F.; Sparks; Ourselin, S. TorchIO: A Python library for efficient loading, preprocessing, augmentation and patch-based sampling of medical images in deep learning. Comput. Methods Programs Biomed. 2021, 208, 106236.

Turinici, G. The convergence of the Stochastic Gradient Descent (SGD): A self-contained proof. arXiv 2021, arXiv:2103.14350.

Chen, J.; Liu, S.; Liu, Y. ALKU-Net: Adaptive Large Kernel Attention Convolution Network for Lung Nodule Segmentation. Electronics 2024, 13, 3121.

Huang, D.; Li, Z.; Jiang, T.; Yang, C.; Li, N. Artificial intelligence in lung cancer: Current applications, future perspectives, and challenges. Front. Oncol. 2024, 14, 1486310.

Frontiers in Medical and Clinical Sciences

Article Details Page