Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2022 Feb 17:5:782225.
doi: 10.3389/frai.2022.782225. eCollection 2022.

Segmentation of Lung Nodules on CT Images Using a Nested Three-Dimensional Fully Connected Convolutional Network

Shoji Kido  1 , Shunske Kidera  2 , Yasushi Hirano  3 , Shingo Mabu  2 , Tohru Kamiya  4 , Nobuyuki Tanaka  5 , Yuki Suzuki  1 , Masahiro Yanagawa  6 , Noriyuki Tomiyama  6
Affiliations

Segmentation of Lung Nodules on CT Images Using a Nested Three-Dimensional Fully Connected Convolutional Network

Shoji Kido et al. Front Artif Intell. .

Abstract

In computer-aided diagnosis systems for lung cancer, segmentation of lung nodules is important for analyzing image features of lung nodules on computed tomography (CT) images and distinguishing malignant nodules from benign ones. However, it is difficult to accurately and robustly segment lung nodules attached to the chest wall or with ground-glass opacities using conventional image processing methods. Therefore, this study aimed to develop a method for robust and accurate three-dimensional (3D) segmentation of lung nodule regions using deep learning. In this study, a nested 3D fully connected convolutional network with residual unit structures was proposed, and designed a new loss function. Compared with annotated images obtained under the guidance of a radiologist, the Dice similarity coefficient (DS) and intersection over union (IoU) were 0.845 ± 0.008 and 0.738 ± 0.011, respectively, for 332 lung nodules (lung adenocarcinoma) obtained from 332 patients. On the other hand, for 3D U-Net and 3D SegNet, the DS was 0.822 ± 0.009 and 0.786 ± 0.011, respectively, and the IoU was 0.711 ± 0.011 and 0.660 ± 0.012, respectively. These results indicate that the proposed method is significantly superior to well-known deep learning models. Moreover, we compared the results obtained from the proposed method with those obtained from conventional image processing methods, watersheds, and graph cuts. The DS and IoU results for the watershed method were 0.628 ± 0.027 and 0.494 ± 0.025, respectively, and those for the graph cut method were 0.566 ± 0.025 and 0.414 ± 0.021, respectively. These results indicate that the proposed method is significantly superior to conventional image processing methods. The proposed method may be useful for accurate and robust segmentation of lung nodules to assist radiologists in the diagnosis of lung nodules such as lung adenocarcinoma on CT images.

Keywords: SegNet; U-Net; computer-aided diagnosis; deep learning; graph cut; lung nodule; segmentation; watershed.

PubMed Disclaimer

Conflict of interest statement

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Figures

Figure 1
Figure 1
Architecture of the proposed nested three-dimensional (3D) fully connected convolutional network. The connections are indicated by the red circles, where the encoder and decoder are connected by concatenation.
Figure 2
Figure 2
Architecture of the residual unit. (A) Conventional feed-forward neural network and (B) residual unit.
Figure 3
Figure 3
Examples of segmentation results in the case of a GGO nodule.
Figure 4
Figure 4
Examples of segmentation results in the case of a nodule attached to the chest wall.
Figure 5
Figure 5
An example of extraction results when the value of λ was changed. Under-extraction was observed when only Dice loss was used as the loss function (λ = 0.0).
Figure 6
Figure 6
An example of extraction results when the value of λ was changed. Over-extraction was observed when only binary cross entropy was used as the loss function (λ = 1.0).
See this image and copyright information in PMC

References

    1. Akiba T., Sano S., Yanase T., Ohta T., Koyama M. (2019). “Optuna: a next-generation hyperparameter optimization framework,” in Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining (Anchorage, AK: ), 2623–2631. 10.1145/3292500.3330701 - DOI
    1. Amorim P. H. J., de Moraes T. F., da Silva J. V. L., Pedrini H. (2019). Lung nodule segmentation based on convolutional neural networks using multi-orientation and patchwise mechanisms. Lect. Not. Comput. Vis. Biomech. 34, 286–295. 10.1007/978-3-030-32040-9_30 - DOI
    1. Armato S. G., McLennan G., Bidaut L., McNitt-Gray M. F., Meyer C. R., Reeves A. P., et al. (2011). The Lung Image Database Consortium (LIDC) and Image Database Resource Initiative (IDRI): a completed reference database of lung nodules on CT scans. Med. Phys. 38, 915–931. 10.1118/1.3528204 - DOI - PMC - PubMed
    1. Badrinarayanan V., Kendall A., Cipolla R. (2017). SegNet: a deep convolutional encoder-decoder architecture for image segmentation. IEEE Trans. Pattern Anal. Machine Intelligence 39, 2481–2495. 10.1109/TPAMI.2016.2644615 - DOI - PubMed
    1. Bertels J., Eelbode T., Berman M., Vandermeulen D., Maes F., Bisschops R., et al. (2019).: Theory and practice. Lect. Not. Comput. Sci. 11765, 92–100. 10.1007/978-3-030-32245-8_11 - DOI

LinkOut - more resources