Classification of lung cancer using deep learning and machine learning in order to increase accuracy in early diagnosis

Subject Areas :

Fattaneh Mohammadi Nezhad ¹ , Mohammad Reza Mollahosseini Ardakani ²

1 - Doctoral student of Computer Engineering, Faculty of Technology and Engineering, Meybod Branch, Islamic Azad University, Meybod, Iran
2 - Department of Computer Engineering, Technical and Engineering Faculty, Meybod Branch, Islamic Azad University, Meybod, Iran

Received: 2024-08-11 Accepted : 2024-10-23 Published : 2024-12-19

Keywords: Feature extraction, lung cancer, support vector machine, deep learning, classification,

Abstract :

Given that lung cancer is one of the most fatal and progressive types of cancer, numerous studies focus on using computer-aided systems for its detection. However, there's a need for further development of intelligent systems to achieve high sensitivity, detection, and accuracy in lung segmentation, particularly for identifying various nodule shapes. This challenge highlights the importance of studying the implementation of an AI-based framework for lung cancer detection. In this research, we utilize the LIDC database, which comprises a documented collection of chest CT scans. The proposed system, based on machine learning and deep learning approaches, includes stages of CT scan image reading, image pre-processing, segmentation, feature extraction, and classification. To prevent the loss of critical features, CT scan images are read directly in their raw DICOM file format. Subsequently, image refinement and enhancement techniques are applied using image processing. For image segmentation and the extraction of its performance metrics, Otsu's method, edge detection, and morphological operations are employed. This study proposes a cancer diagnostic model based on an active machine learning approach combined with a deep learning technique. The designed computer-aided diagnostic (CAD) model accurately identifies physiological and pathological changes in the soft tissues of lung cancer lesions, even at early stages of the disease. The results demonstrate that the proposed approach, by combining the advantages of a) the diagnostic accuracy of deep learning methods and b) the simplicity of machine learning classification, can significantly assist radiologists in the early detection of lung cancer and facilitate timely patient management

References:

[1] M. Islam, R. Rab, “Analysis of CT scan images to predict lung cancer stages using image processing techniques”, in 2019 IEEE 10th Annual Information Technology, 2019, pp. 943-953.
[2] A. Singareddy, M. E. Flanagan, P. P. Samson, S. N. Waqar, “Trends in stage I lung cancer”, Clinical Lung Cancer, vol. 31, no. 9, pp. 114-119, 2023.
[3] M. S. Islam, R. Islam, “A Critical Survey on Developed Reconstruction Algorithms for Computed Tomography Imaging from a Limited Number of Projections”, International Journal of Image and Graphics, vol. 23, no. 4, pp. 235-243, 2023.
[4] T. Nakra, A. Mehta, A. Bal, A. Nambirajan, D. Mishra, D, P. Midha, P. Gupta, et al. ”Epidermal growth factor receptor mutation status in pulmonary adenocarcinoma: Multi-institutional data discussion at national conference of “Lung Cancer Management in Indian context””, Current Problems in Cancer, vol. 44, no. 3, pp. 751-762, Sep. 2020.
[5] T. Pereira, C. Freitas, J.L. Costa, J. Morgado, F. Silva, E. Negrão, B.F. de Lima et al., “Comprehensive Perspective for Lung Cancer Characterisation Based on AI Solutions Using CT Images”, Journal of Clinical Medicine, vol. 10, no. 1, pp. 339-346, 2020.
[6] W. J. Sori, J. Feng, A. W. Godana, S. Liu, “DFD-Net: lung cancer detection from denoised CT scan image using deep learning”, Frontiers of Computer Science, vol. 15, no. 3, pp. 425-438, 2021.
[7] F. Bray, J. Ferlay, I. Soerjomataram, R. L. Siegel, L. A. Torre, A. Jemal, “Global cancer statistics 2018”, A Cancer Journal for Clinicians, vol, 68, no. 6, pp. 394–424, 2018.
[8] M. Reck, S. Dettmer, H. U. Kauczor, R. Kaaks, “Lung cancer screening with low-dose computed tomography: current status in Germany”, Deutsches Ärzteblatt International, vol. 120, no. 23, pp. 387-392, 2023.
[9] L. Guo, Y. Yu, F. Yang, W. Gao, “Accuracy of baseline low-dose computed tomography lung cancer screening: a systematic review and meta-analysis”, Chinese Medical Journal, vol. 136, no. 9, pp. 1047-1056, 2023.
[10] S. Üzülmez, M. A. Çifçi, “Early diagnosis of lung cancer using deep learning and uncertainty measures”, Journal of the Faculty of Engineering and Architecture of Gazi University, vol. 39, no. 1, pp. 385-400, 2024.
[11] I. Shafi, S. Din, A. Khan, I. D. L. T. Díez, “An Effective Method for Lung Cancer Diagnosis from CT Scan Using Deep Learning-Based Support Vector Network”, Journal of the Faculty of Engineering, Süleyman Demirel University, vol. 14, no. 1, pp. 1-18, 2022.
[12] R. Mothkur, B. N. Veerappa, “Classification of lung cancer using lightweight deep neural networks”, Procedia Computer Science, vol. 71, no. 1, pp. 1869-1877, 2023.
[13] N. S. Reddy and V. Khanaa, “Intelligent deep learning algorithm for lung cancer detection and classification”, Bulletin of Electrical Engineering and Informatics, vol. 35, no. 3, pp. 2089-2096, 2023.
[14] D. Moitra and R. K. Mandal, “Classification of non-small cell lung cancer using one-dimensional convolutional neural network”, Expert Systems with Applications, vol. 30, no. 2, pp. 852-863, 2020.
[15] Z. Yu, P. Ni, H. Yu, T. Zuo, and Y. Liu, “Effectiveness of a single low-dose computed tomography screening for lung cancer: A population-based perspective cohort study in China”, International Journal of Cancer, vol. 154, no. 4, pp. 659-669, 2024.
[16] M. T. H. Chowdhury, “Application of Signal Processing and Deep Hybrid Learning in Phonocardiogram and Electrocardiogram Signals to Detect Early Stage Heart Diseases”, M.Sc. dissertation, June 2022.
[17] G. A. H. Alabandi, “Combining Deep Learning with Traditional Machine Learning to Improve Classification Accuracy”, M.Sc. dissertation, Department of Computer Science, Texas State University, 2017.
[18] A. Bhattacharya, “Deep Hybrid Learning — a fusion of conventional ML with state of the art DL”, in GIDS.AI/ML and Data Live, 2020, pp. 425-437.
[19] D. Sengupta et al., “A deep hybrid learning pipeline for accurate diagnosis of ovarian cancer based on nuclear morphology”, PLOS ONE, vol. 17, no. 1, pp. 1-20, Jan. 2022.
[20] M. Islam, R. Rab, “Analysis of CT scan images to predict lung cancer stages using image processing techniques”, in 2019 IEEE 10th Annual Information Technology Conference (ITech), 2019, pp. 961-967