A combined automated diagnostic system for segmentation and detection of lung cancer nodules

A combined automated diagnostic system for segmentation and detection of lung cancer nodules

Anna A. Meldo
PhD in Medical Sciences, Saint-Petersburg Clinical Research Center of Specialized Types of Medical Care (Oncological), Head of the Radiology Department, Radiologyst, 68A, Leningradskaja ul., Pesochny, Leningrad region, 197758, Russia, Peter the Great Saint-Petersburg Polytechnical University (SPbPU), Research Laboratory of Neural Network Technol-ogies and Artificial Intelligence, Senior Research Scientist, 29, Politekhnich-eskaya ul., Saint-Petersburg, 195251, Russia, This email address is being protected from spambots. You need JavaScript enabled to view it.

Lev V. Utkin
Doctor of Technical Science, Peter the Great Saint-Petersburg Polytechnical University (SPbPU), Research Laboratory of Neural Network Technologies and Artificial Intelligence, Professor, Head of Laboratory, Head of the Telematics Department, 29, Politekhnicheskaya ul., Saint-Petersburg, 195251, Russia, ORCID: 0000-0002-5637-1420, This email address is being protected from spambots. You need JavaScript enabled to view it.

Mikhail A. Ryabinin
Peter the Great Saint-Petersburg Polytechnical University (SPbPU), Research Laboratory of Neural Network Technologies and Artificial Intelligence, Engineer, Post-graduate Student of the Telematics Department, 29, Politekhnicheskaya ul., Saint-Petersburg, 195251, Russia, This email address is being protected from spambots. You need JavaScript enabled to view it.


Received 24 April 2019

Abstract
The paper deals with a problem of automating and intellectualizing a diagnostic study of lung cancer based on computed tomography scans. The relevance of the developed system is based on the fact that lung cancer in many countries is the most common cancer and its detection at an early stage may significantly increase the chance of patient survival. The use of automated diagnostic systems partially solves the problem of diagnosis at an early stage. The article proposes a new architecture of a part of an automated system, the purpose of which is segmentation and detection of nodules in lungs and minimization of missed nodules, which determines the sensitivity of the system. Its peculiarity lies in the fact that, along with deep learning algorithms based on the use of deep neural networks (2D U-Net and 3D U-Net), traditional computed tomography image processing algorithms are used in parallel, which increase the system's ability to detect tumors. The system is a three-channel data processing system, where each channel is a procedure for detecting nodules. The training process of the system is based on the use of an open labelled database of computed tomography scans of patients called LU-NA16 and a new database containing atypical cases of lung cancer called LIRA. The implementation of the preprocessing procedure and of the lung segmentation procedure are considered. Architectures of deep neural networks are proposed which are the most suitable for detecting nodules. Segmented nodules of a new patient can be viewed as the system output.

Key words
Deep neural network, image processing, image segmentation, lung cancer, artificial intelligence, computed tomography.

Acknowledgements
This work is supported by the Russian Science Foundation under grant 18-11-00078.

DOI

https://doi.org/10.31776/RTCJ.7209

Bibliographic description
Meldo, A., Utkin, L. and Ryabinin, M. (2019). A combined automated diagnostic system for segmentation and detection of lung cancer nodules. Robotics and Technical Cybernetics, 7(2), pp.145-153.

UDC identifier:
004.896

References

  1. Siegel, R., Miller, K. and Jemal, A. (2018). Cancer statistics, 2018. CA: A Cancer Journal for Clinicians, 68(1), pp.7-30.
  2. Kang, G., Liu, K., Hou, B. and Zhang, N. (2017). 3D multi-view convolutional neural networks for lung nodule classification. PLOS ONE, 12(11), p.e0188290.
  3. Choi, W. and Choi, T. (2014). Automated pulmonary nodule detection based on three-dimensional shape-based feature descriptor. Computer Methods and Programs in Biomedicine, 113(1), pp.37-54.
  4. Shaukat, F., Raja, G., Gooya, A. and Frangi, A. (2017). Fully automatic detection of lung nodules in CT images using a hybrid feature set. Medical Physics, 44(7), pp.3615-3629.
  5. Zhang, G., Jiang, S., Yang, Z., Gong, L., Ma, X., Zhou, Z., Bao, C. and Liu, Q. (2018). Automatic nodule detection for lung cancer in CT images: A review. Computers in Biology and Medicine, 103, pp.287-300.
  6. Zhang, J., Xia, Y., Cui, H. and Zhang, Y. (2018). Pulmonary nodule detection in medical images: A survey. Biomedical Signal Processing and Control, 43, pp.138-147.
  7. Litjens, G., Kooi, T., Bejnordi, B., Setio, A., Ciompi, F., Ghafoorian, M., van der Laak, J., van Ginneken, B. and Sánchez, C. (2017). A survey on deep learning in medical image analysis. Medical Image Analysis, 42, pp.60-88.
  8. Setio, A. et al. (2017). Validation, comparison, and combination of algorithms for automatic detection of pulmonary nodules in computed tomography images: The LUNA16 challenge. Medical Image Analysis, 42, pp.1-13.
  9. Çiçek, Ö., Abdulkadir, A., Lienkamp, S., Brox, T. and Ronneberger, O. (2016). 3D U-Net: Learning Dense Volumetric Segmentation from Sparse Annotation. Medical Image Computing and Computer-Assisted Intervention – MICCAI 2016, pp.424-432.
  10. Ronneberger, O., Fischer, P. and Brox, T. (2015). U-Net: Convolutional Networks for Bio-medical Image Segmentation. Lecture Notes in Computer Science, pp.234-241.
  11. Armato, S. et al. (2011). The Lung Image Database Consortium (LIDC) and Image Data-base Resource Initiative (IDRI): A Completed Reference Database of Lung Nodules on CT Scans. Medical Physics, 38(2), pp.915-931.
  12. Utkin, L., Ryabinin, M., Meldo, A., Lukashin, A. and Prokhorov, I. (2019). Baza dannih computernih tomogramm grudnoy kletki s videlannimi i markirovannimi oblastami pathology legkih – LIRA (Lung Image Resource Annotated). 2019620232 (in Russ.).
  13. Firmino, M., Morais, A., Mendoça, R., Dantas, M., Hekis, H. and Valentim, R. (2014). Computer-aided detection system for lung cancer in computed tomography scans: Review and future prospects. BioMedical Engineering OnLine, 13(1), p.41.
  14. Ozdemir, O., Russell, R. and Berlin, A. (2019). A 3D Probabilistic Deep Learning System for Detection and Diagnosis of Lung Cancer Using Low-Dose CT Scans. [online] Export.arxiv.org. Available at: http://export.arxiv.org/abs/1902.03233v1 [Accessed 24 Apr. 2019].
  15. Firmino, M., Angelo, G., Morais, H., Dantas, M. and Valentim, R. (2016). Computer-aided detection (CADe) and diagnosis (CADx) system for lung cancer with likelihood of malignancy. BioMedical Engineering OnLine, 15(1).
  16. Lowekamp, B., Chen, D., Ibáñez, L. and Blezek, D. (2013). The Design of SimpleITK. Frontiers in Neuroinformatics, 7.
  17. Lehmann, G. (2008). Label object representation and manipulation with ITK. The Insight Journal, pp.1-34.
  18. Teramoto, A. and Fujita, H. (2012). Fast lung nodule detection in chest CT images using cylindrical nodule-enhancement filter. International Journal of Computer Assisted Radiology and Surgery, 8(2), pp.193-205.
Editorial office address: 21, Tikhoretsky pr., Saint-Petersburg, Russia, 194064, tel.: +7(812) 552-13-25 e-mail: zheleznyakov@rtc.ru