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
. 2021 Nov 8;22(Suppl 5):314.
doi: 10.1186/s12859-021-04234-0.

Design of lung nodules segmentation and recognition algorithm based on deep learning

Hui Yu  1 , Jinqiu Li  1 , Lixin Zhang  1 , Yuzhen Cao  1 , Xuyao Yu  2 , Jinglai Sun  3
Affiliations

Design of lung nodules segmentation and recognition algorithm based on deep learning

Hui Yu et al. BMC Bioinformatics. .

Abstract

Background: Accurate segmentation and recognition algorithm of lung nodules has great important value of reference for early diagnosis of lung cancer. An algorithm is proposed for 3D CT sequence images in this paper based on 3D Res U-Net segmentation network and 3D ResNet50 classification network. The common convolutional layers in encoding and decoding paths of U-Net are replaced by residual units while the loss function is changed to Dice loss after using cross entropy loss to accelerate network convergence. Since the lung nodules are small and rich in 3D information, the ResNet50 is improved by replacing the 2D convolutional layers with 3D convolutional layers and reducing the sizes of some convolution kernels, 3D ResNet50 network is obtained for the diagnosis of benign and malignant lung nodules.

Results: 3D Res U-Net was trained and tested on 1044 CT subcases in the LIDC-IDRI database. The segmentation result shows that the Dice coefficient of 3D Res U-Net is above 0.8 for the segmentation of lung nodules larger than 10 mm in diameter. 3D ResNet50 was trained and tested on 2960 lung nodules in the LIDC-IDRI database. The classification result shows that the diagnostic accuracy of 3D ResNet50 is 87.3% and AUC is 0.907.

Conclusion: The 3D Res U-Net module improves segmentation performance significantly with the comparison of 3D U-Net model based on residual learning mechanism. 3D Res U-Net can identify small nodules more effectively and improve its segmentation accuracy for large nodules. Compared with the original network, the classification performance of 3D ResNet50 is significantly improved, especially for small benign nodules.

Keywords: Convolutional neural network; Image classification; Image segmentation; Lung nodule; Residual learning; U-Net.

PubMed Disclaimer

Conflict of interest statement

The authors declare that they have no competing interests.

Figures

Fig. 1
Fig. 1
Overview of lung nodule detection and diagnosis system
Fig. 2
Fig. 2
The process of segmenting lung parenchyma from CT images
Fig. 3
Fig. 3
The architecture of 3D Res U-Net
Fig. 4
Fig. 4
Residual unit
Fig. 5
Fig. 5
The architecture of 3D ResNet50
Fig. 6
Fig. 6
The loss of 3D Res U-Net and 3D U-Net changes during the training process. The loss function used by 3D U-Net is Dice Loss, and 3D Res U-Net used a mixed loss function
Fig. 7
Fig. 7
The lung nodule segmentation results using 3D U-Net and 3D Res U-Net
Fig. 8
Fig. 8
Training loss and validation loss of 3D ResNet50
Fig. 9
Fig. 9
Confusion matrices of classification results using different networks and ROC curve of classification results using 3D ResNet50
Fig. 10
Fig. 10
Segmentation results of lung nodules with different sizes by 3D Res U-Net. a Lung mass. b Nodule. c Small nodule. d Micro nodule
Fig. 11
Fig. 11
Some malignant ground glass opacity nodules that are misclassified as benign nodules and the probabilities of benign and malignant nodules made by 3D ResNet50
See this image and copyright information in PMC

References

    1. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2019. CA Cancer J Clin. 2019;69(1):7–34. - PubMed
    1. de Carvalho Filho AO, de Sampaio WB, Silva AC, de Paiva AC, Nunes RA, Gattass M. Automatic detection of solitary lung nodules using quality threshold clustering, genetic algorithm and diversity index. Artif Intell Med. 2014;60(3):165–177. doi: 10.1016/j.artmed.2013.11.002. - DOI - PubMed
    1. Jacobs C, van Rikxoort EM, Twellmann T, Scholten ET, de Jong PA, Kuhnigk JM, et al. Automatic detection of subsolid pulmonary nodules in thoracic computed tomography images. Med Image Anal. 2014;18(2):374–384. doi: 10.1016/j.media.2013.12.001. - DOI - PubMed
    1. Li B, Chen K, Peng G, Guo Y, Tian L, Ou S, et al. Segmentation of ground glass opacity pulmonary nodules using an integrated active contour model with wavelet energy-based adaptive local energy and posterior probability-based speed function. Mater Express. 2016;6(4):317–327. doi: 10.1166/mex.2016.1311. - DOI
    1. Mao Q, Zhao S, Gong T, Zheng Q. An Effective Hybrid Windowed Fourier Filtering and Fuzzy C-Mean for Pulmonary Nodule Segmentation. J Med Imaging Health Inf. 2018;8(1):72–77. doi: 10.1166/jmihi.2018.2235. - DOI