Automated recognition of lung diseases in CT images based on the optimum-path forest classifier
The World Health Organization estimated that around 300 million people have asthma, and 210 million people are affected by Chronic Obstructive Pulmonary Disease (COPD). Also, it is estimated that the number of deaths from COPD increased \(30\%\) in 2015 and COPD will become the third major cause of death worldwide by 2030. These statistics about lung diseases get worse when one considers fibrosis, calcifications and other diseases. For the public health system, the early and accurate diagnosis of any pulmonary disease is mandatory for effective treatments and prevention of further deaths. In this sense, this work consists in using information from lung images to identify and classify lung diseases. Two steps are required to achieve these goals: automatically extraction of representative image features of the lungs and recognition of the possible disease using a computational classifier. As to the first step, this work proposes an approach that combines Spatial Interdependence Matrix (SIM) and Visual Information Fidelity (VIF). Concerning the second step, we propose to employ a Gaussian-based distance to be used together with the optimum-path forest (OPF) classifier to classify the lungs under study as normal or with fibrosis, or even affected by COPD. Moreover, to confirm the robustness of OPF in this classification problem, we also considered Support Vector Machines and a Multilayer Perceptron Neural Network for comparison purposes. Overall, the results confirmed the good performance of the OPF configured with the Gaussian distance when applied to SIM- and VIF-based features. The performance scores achieved by the OPF classifier were as follows: average accuracy of \(98.2\%\), total processing time of 117 microseconds in a common personal laptop, and F-score of 95.2% for the three classification classes. These results showed that OPF is a very competitive classifier, and suitable to be used for lung disease classification.
KeywordsMedical imaging Optimum-path forest Feature extraction Image classification
The authors thank the Graduate Program in Computer Science from the Federal Institute of Education, Science and Technology of Ceará and the Department of Computer Engineering from the Walter Cantídio University Hospital of the Federal University of Ceará, in Brazil, for the support given.
The first author acknowledges the sponsorship from the Federal Institute of Education, Science and Technology of Ceará through grants PROINFRA/2013 and PROAPP/2014. The author acknowledges also the sponsorship from the Brazilian National Council for Research and Development (CNPq).
Victor Hugo C. de Albuquerque thanks CNPq for providing financial support through grants 470501/2013-8 and 301928/2014-2.
João P. Papa is grateful to São Paulo Research Foundation grants #2014/16250-9 and #2014/12236-1, as well as CNPq grant #306166/2014-3.
Authors gratefully acknowledge the funding of Project NORTE-01-0145-FEDER-000022—SciTech—Science and Technology for Competitive and Sustainable Industries, cofinanced by “Programa Operacional Regional do Norte” (NORTE2020), through “Fundo Europeu de Desenvolvimento Regional” (FEDER).
Compliance with ethical standards
Conflict of interest
The authors report no conflict of interest.
- 1.WHO (2016) Causes of death in the world. Technical report, World Health OrganizationGoogle Scholar
- 5.WHO (2016) Chronic obstructive pulmonary disease (copd). Technical report, World Health OrganizationGoogle Scholar
- 8.Rebouças Filho PP, Cortez PC, da Silveira Tarique Felix JHS, Cavalcante TS, Holanda MA (2013) Adaptive 2D crisp active contour model applied to lung segmentation in CT images of the thorax of healthy volunteers and patients with pulmonary emphysema. Revista Brasileira de Engenharia Biomédica 29:363–376CrossRefGoogle Scholar
- 11.Liang T K, Tanaka T, Nakamura H, Shirahata T, Sugiura H (2008) An automated 3D emphysema extraction method using lung CT. In: SICE Annual Conference 2008, pp 3110–3114Google Scholar
- 14.Ma Z, Tavares JMRS, Jorge RMN (2009) A review on the current segmentation algorithms for medical images. In: 1st international conference on imaging theory and applications (IMAGAPP) 5(8):135–140Google Scholar
- 17.Oliveira M Costa, Ferreira J Raniery (2013) A bag-of-tasks approach to speed up the lung nodules retrieval in the bigdata age. In: IEEE 15th international conference o e-health networking, applications services (Healthcom), pp 632–636Google Scholar
- 25.Suzuki C, Gomes J, Falcão A, Papa JP, Hoshino-Shimizu S (2012) Automatic segmentation and classification of human intestinal parasites from microscopy images. IEEE Trans Biomed Eng 60(9):803–812Google Scholar
- 26.Capabianco FAM, Falcão AX, Yasuda CL, Udupa JK (2012) Brain tissue MR-image segmentation via optimum-path forest clustering. IEEE Trans Image Process 116:1047–1059Google Scholar
- 33.Nunes TM, de Albuquerque Victor Hugo C, Papa JP, Silva CC, Normando Paulo G, Moura Elineudo P, Tavares João Manuel RS (2013) Automatic microstructural characterization and classification using artificial intelligence techniques on ultrasound signals. Expert Syst Appl 40:3096–3105CrossRefGoogle Scholar
- 34.Gomes Samuel L, Rebouças Elizângela de S, Neto Edson Cavalcanti, Papa João P, de Albuquerque Victor HC, Rebouças Filho Pedro P, Tavares João Manuel RS (2016) Embedded real-time speed limit sign recognition using image processing and machine learning techniques. Neural Comput Appl 1–12. doi: 10.1007/s00521-016-2388-3
- 35.Silva EM, Marinho LB, Rebouças Filho PP, Leite JP, Leite Josinaldo P, Fialho Walter M L, de Albuquerque Victor Hugo C, Tavares João Manuel R S (2016) Classification of induced magnetic field signals for the microstructural characterization of sigma phase in duplex stainless steels. Metals 6(7):164CrossRefGoogle Scholar
- 37.de Albuquerque VHC, Barbosa CV, Silva CC, Moura EP, Rebouças Filho Pedro P, Papa João P, Tavares João Manuel R S (2015) Ultrasonic sensor signals and optimum path forest classifier for the microstructural characterization of thermally-aged inconel 625 alloy. Sensors 15(6):12474CrossRefGoogle Scholar
- 41.Turesson Hjalmar K, Ribeiro S, Pereira DR, Papa JP, de Albuquerque VHC (2016) Machine learning algorithms for automatic classification of marmoset vocalizations. PLoS ONE 11(9):1–14Google Scholar
- 42.de Albuquerque VHC, Nunes TM, Pereira DR, Luz EJDS, Menotti D, Papa João P, Tavares João Manuel RS (2016) Robust automated cardiac arrhythmia detection in ECG beat signals. Neural Comput Appl 1–15. doi: 10.1007/s00521-016-2472-8
- 45.De Alexandria AR, Cortez PC, Bessa JA, da Silva Félix JH, De Abreu José Sebastião, De Albuquerque Victor Hugo C (2014) psnakes: A new radial active contour model and its application in the segmentation of the left ventricle from echocardiographic images. Comput Methods Programs Biomed 116(3):260–273CrossRefGoogle Scholar
- 47.Moreira FDL, Kleinberg MN, Arruda HF, Freitas FNC, Parente Marcelo Monteiro Valente, de Albuquerque Victor Hugo Costa, Rebouças Filho Pedro Pedrosa (2016) A novel vickers hardness measurement technique based on adaptive balloon active contour method. Expert Syst Appl 45:294–306CrossRefGoogle Scholar
- 48.Rebouças Pedro Pedrosa, Sarmento Roger Moura, Cortez Paulo C, Antˆonio Carlos Da Silva Barros, De Albuquerque Victor Hugo C (2015) Adaptive crisp active contour method for segmentation and reconstruction of 3d lung structures. Int J Comput Appl 111(4):1–8Google Scholar
- 49.Neto EC, Cortez PC, Cavalcante TS, da Silva Filho VER, Rebouças Filho PP, Holanda MA (2015) Supervised enhancement filter applied to fissure detection. Springer, Cham, pp 337–340Google Scholar
- 51.Valente IRS, Cortez PC, Neto EC, Soares JM, de Albuquerque Victor Hugo C, Tavares João Manuel RS (2016) Automatic 3D pulmonary nodule detection in ct images: a survey. Comput Methods Programs Biomed 124:91–107Google Scholar
- 52.Papa JP, Suzuki CTN, Falcão AX (2009) LibOPF: A library for the design of Optimum-Path Forest classifiers, Software version 2.0 available at http://www.ic.unicamp.br/~afalcao/LibOPF
- 53.Felix JHS, Cortez PC, Holanda MA, Costa RCS (2007) Automatic segmentation and measurement of the lungs in healthy persons and in patients with chronic obstructive pulmonary disease in CT images. vol. 18, pp 370–373, Margarita Island, Venezuela, October 2007. In: IV Latin American Congress on Biomedical Engineering 2007, Bioengineering Solutions for Latin America HealthGoogle Scholar
- 54.Felix John Heber S, Cortez PC, Holanda MA, Colaço DF, Albuquerque VHC, Alexandria AR (2007) Lung and chest wall structures segmentation in CT images. pp 291–294. In: Computational Vision and Medical Image Processing (VIPMAGE)Google Scholar
- 57.Rebouças Filho Pedro P, Rebouças Elizângela de S, Marinho Leandro B, Sarmento Róger M, Tavares João Manuel RS, de Albuquerque Victor Hugo C (2017) Analysis of human tissue densities: a new approach to extract features from medical images. Pattern Recogn Lett. doi: 10.1016/j.patrec.2017.02.005
- 59.Medeiros Fátima NS, Ramalho Geraldo LB, Bento Mariana P, Medeiros Luiz CL (2010) On the evaluation of texture and color features for nondestructive corrosion detection. EURASIP J Adv Signal Process 2010:7Google Scholar
- 60.Ultsch A (2013) U * -matrix: a tool to visualize clusters in high dimensional data, vol 36. University of Marburg, Department of Computer Science, MarburgGoogle Scholar
- 63.Allène C, Audibert JY, Couprie M, Cousty J, Keriven R (2007) Some links between min-cuts, optimal spanning forests and watersheds. In: Proceedings of the 8th international symposium on mathematical morphology, pp 253–264Google Scholar
- 64.Liu H, Jiang S, Huang Q, Xu C, Gao W (2007) region-based visual attention analysis with its application in image browsing on small displays. In: Proceedings of the 15th international conference on multimedia, pp25–29Google Scholar
- 66.Nissen S (2013) Implementation of a fast artificial neural network library (FANN), 2003. Department of Computer Science University of Copenhagen (DIKU). Software available at http://leenissen.dk/fann/
- 67.Haykin SO (2008) Neural networks and learning machines. Pearson Prentice Hall, Upper Saddle RiverGoogle Scholar
- 71.Siegel S (2006) Estatística não-paramétrica para as ciências do comportamento. Série Métodos de Pesquisa, vol 2. Artmed, Porto Alegre, p 7Google Scholar
- 72.Triola Mario F et al (2005) Introdução à estatística, vol 10. LTC, Rio de JaneiroGoogle Scholar