Transactions on Machine Intelligence

Transactions on Machine Intelligence

A New Approach to Feature Extraction Based on Lung CT Images Using Machine Learning Algorithms for Lung Disease Classification

Document Type : Original Article

Authors
M. R. Fazel Najafabadi
S. Ayat Najafabadi
S. Fekri Ershad
Department of Computer Engineering, Najafabad Branch, Islamic Azad University, Isfahan, Iran
10.47176/TMI.2018.99
Abstract
Accurate diagnosis of lung diseases based on processing and analyzing lung CT images is crucial for aiding medical decision-making. This study presents a new feature extraction method based on human tissue density patterns, called Analysis of Human Tissue Density (AHTD). This method is compared with the Gray Level Co-occurrence Matrix (GLCM), Hu Moments (HM), Statistical Moments (SM), and Zernike Moments (ZM). The dataset of chest tomography images was obtained from the Walter Cantidio University Hospital in Fortaleza, Brazil. Four machine learning classifiers were used in this study: Bayesian Classifier, Optimum-Path Forest (OPF), k-Nearest Neighbors (KNN), and Support Vector Machine (SVM) to classify lung diseases in chest images. Feature extraction from lung images was performed in 5.2 milliseconds, achieving an accuracy of 99.01% for lung disease diagnosis and classification. The results of this study suggest that the proposed method can be used in real-time applications due to its rapid processing time and high accuracy for classifying lung diseases based on lung CT images.
Keywords
Human Tissue Density Analysis
Gray Level Co-occurrence Matrix
Lung Disease
Moments
Machine Learning
Feature Extraction
Support Vector Machine
Optimum-Path Forest
[1]    Khan, I. Y., Zope, P. H., & Suralkar, S. R. (2013). Importance of artificial neural network in medical diagnosis of diseases like acute nephritis disease and heart disease. International Journal of Engineering Science and Innovative Technology, 2(2), 210-217.
[2]    Temurtas, H., Yumusak, N., & Temurtas, F. (2009). A comparative study on diabetes disease diagnosis using neural networks. Expert Systems with Applications, 36(4), 8610-8615. https://doi.org/10.1016/j.eswa.2008.10.032
[3]    Ozkan, H., Osman, O., & Sahin, S. (2013). Computer aided detection of pulmonary embolism in computed tomography angiography images. In 2013 International Conference on Electronics, Computer and Computation (ICECCO) (pp. 355-358). https://doi.org/10.1109/ICECCO.2013.6718301
[4]    Rebouças Filho, P. P., de S. Rebouças, E., Marinho, L. B., Sarmento, R. M., Tavares, J. M. R. S., & de Albuquerque, V. H. C. (2017). Analysis of human tissue densities: A new approach to extract features from medical images. Pattern Recognition Letters. https://doi.org/10.1016/j.patrec.2017.02.005
[5]    Ramalho, G. L. B., Rebouças Filho, P. P., de Medeiros, F. N. S., & Cortez, P. C. (2014). Lung disease detection using feature extraction and extreme learning machine. Revista Brasileira de Engenharia Biomédica, 30(3), 207-214. https://doi.org/10.1590/rbeb.2014.019
[6]    Neto, E. C., Cortez, P. C., Cavalcante, T. S., Rodrigues, V. E., Reboucas Filho, P. P., & Holanda, M. A. (2016). 3D lung fissure segmentation in TC images based in textures. IEEE Latin America Transactions, 14(1), 254-258. https://doi.org/10.1109/TLA.2016.7430087
[7]    Pforte, A. (2004). Epidemiology, diagnosis, and therapy of pulmonary embolism. European Journal of Medical Research, 9(4), 171-179.
[8]    Eskildsen, S. F., Coupé, P., Fonov, V. S., Pruessner, J. C., & Collins, D. L. (2015). Structural imaging biomarkers of Alzheimer's disease: Predicting disease progression. Neurobiology of Aging, 36(S1), S23-S31. https://doi.org/10.1016/j.neurobiolaging.2014.04.034
[9]    Gonzalez, R. C., & Woods, R. E. (1992). Digital image processing. Addison-Wesley.
[10]    Khotanzad, A., & Hong, Y. (1990). Invariant image recognition by Zernike moments. IEEE Transactions on Pattern Analysis and Machine Intelligence, 12(5), 489-497. https://doi.org/10.1109/34.55109
[11]    Teh, C.-H., & Chin, R. T. (1988). On image analysis by the methods of moments. IEEE Transactions on Pattern Analysis and Machine Intelligence, 10(4), 496-513. https://doi.org/10.1109/34.3913
[12]    Reboucas, F., Pedro, P., Cortez, P. C., & Holanda, M. A. (2011). Active contour models CRISP: New technique for segmentation of the lungs in CT images. Revista Brasileira de Engenharia Biomédica, 27(4), 259-272. https://doi.org/10.4322/rbeb.2011.021
[13]    Jurkovic, I.-A., Stathakis, S., Papanikolaou, N., & Mavroidis, P. (2016). Prediction of lung tumor motion extent through artificial neural network (ANN) using tumor size and location data. Biomedical Physics & Engineering Express, 2(2), 025012. https://doi.org/10.1088/2057-1976/2/2/025012
Transactions on Machine Intelligence
Volume 1, Issue 2
Spring 2018
Pages 99-105

  • Receive Date 10 March 2024
  • Revise Date 24 May 2024
  • Accept Date 26 June 2024