Automated system for lung nodules classification based on wavelet feature descriptor and support vector machine

BioMedical Engineering OnLine

Manuel de Jesús Nandayapa Alfaro²  Humberto de Jesús Ochoa Domínguez¹  Vianey Guadalupe Cruz Sánchez¹  Osslan Osiris Vergara Villegas²  Hiram Madero Orozco² 

[1]Departamento de Ingeniería Eléctrica y Computación, Instituto de Ingeniería y Tecnología, Universidad Autónoma de Ciudad Juárez, Av. del Charro 450 norte, Ciudad Juárez, Z. C. 32310, Chihuahua, México

[2]Departamento de Ingeniería Industrial y Manufactura, Instituto de Ingeniería y Tecnología, Universidad Autónoma de Ciudad Juárez, Av. del Charro 450 norte, Ciudad Juárez, Z. C. 32310, Chihuahua, México

关键词: Texture;    Support vector machine;    Gray level co-ocurrence matrix;    Wavelet feature descriptor;    CT scan;    Lung nodules;    CADx system;   

Others  :  1127599
DOI  :  10.1186/s12938-015-0003-y

 received in 2014-09-09, accepted in 2015-01-23,  发布年份 2015

【 摘 要 】

Background

Lung cancer is a leading cause of death worldwide; it refers to the uncontrolled growth of abnormal cells in the lung. A computed tomography (CT) scan of the thorax is the most sensitive method for detecting cancerous lung nodules. A lung nodule is a round lesion which can be either non-cancerous or cancerous. In the CT, the lung cancer is observed as round white shadow nodules. The possibility to obtain a manually accurate interpretation from CT scans demands a big effort by the radiologist and might be a fatiguing process. Therefore, the design of a computer-aided diagnosis (CADx) system would be helpful as a second opinion tool.

Methods

The stages of the proposed CADx are: a supervised extraction of the region of interest to eliminate the shape differences among CT images. The Daubechies db1, db2, and db4 wavelet transforms are computed with one and two levels of decomposition. After that, 19 features are computed from each wavelet sub-band. Then, the sub-band and attribute selection is performed. As a result, 11 features are selected and combined in pairs as inputs to the support vector machine (SVM), which is used to distinguish CT images containing cancerous nodules from those not containing nodules.

Results

The clinical data set used for experiments consists of 45 CT scans from ELCAP and LIDC. For the training stage 61 CT images were used (36 with cancerous lung nodules and 25 without lung nodules). The system performance was tested with 45 CT scans (23 CT scans with lung nodules and 22 without nodules), different from that used for training. The results obtained show that the methodology successfully classifies cancerous nodules with a diameter from 2 mm to 30 mm. The total preciseness obtained was 82%; the sensitivity was 90.90%, whereas the specificity was 73.91%.

Conclusions

The CADx system presented is competitive with other literature systems in terms of sensitivity. The system reduces the complexity of classification by not performing the typical segmentation stage of most CADx systems. Additionally, the novelty of the algorithm is the use of a wavelet feature descriptor.

【 授权许可】

   
2015 Madero Orozco et al.; licensee BioMed Central.

【 图 表 】

Figure 1.

Figure 2.

Figure 3.

Figure 4.

Figure 5.

Figure 6.

Figure 7.

Figure 8.

【 参考文献 】

• [1]Iranifam M: Analytical applications of chemiluminescence methods for cancer detection and therapy. Trac–Trend Anal Chem 2014, 59:156-83.
• [2]Siegel R, Ma J, Zou Z, Jemal A: Cancer statistics, 2014. Ca-Cancer J Clin 2014, 64:9-29.
• [3]Levin D, Parker L, Halpern E, Rao V: Are combined CT scans of the thorax being overused? J Am Coll Radiol 2014, 11:788-90.
• [4]Rasmussen J, Siersma V, Pedersen J, Heleno B, Saghir Z, Brodersen J: Healthcare costs in the Danish randomised controlled lung cancer CT-screening trial: a registry study. Lung Cancer 2014, 83:347-55.
• [5]Diciotti S, Lombardo S, Falchini M, Picozzi G, Mascalchi M: Automated segmentation refinement of small lung nodules in CT scans by local shape analysis. IEEE T Bio-Med Eng 2011, 58:3418-28.
• [6]Farag A, El Munim H, Graham J, Farag A: A novel approach for lung nodules segmentation in chest CT using level sets. IEEE T Image Process 2013, 22:5202-13.
• [7]Choi W, Choi T: Automated pulmonary nodule detection based on three-dimensional shape-based feature descriptor. Comput Meth Prog Bio 2014, 113:37-54.
• [8]Firmino M, Morais A, Mendoca R, Dantas M, Hekis H: Computer-aided detection system for lung cancer in computed tomography scans: review and future prospects. Biomed Eng Online 2014, 13:1-16. BioMed Central Full Text
• [9]Li Q: Recent progress in computer-aided diagnosis of lung nodules on thin-section CT. Comput Med Imag Grap 2007, 31:248-57.
• [10]Ambrosini V, Nicolini S, Caroli P, Nanni C, Massaro A, Marzola M, et al.: PET/CT imaging in different types of lung cancer: an overview. Eur J Radiol 2012, 81:988-1001.
• [11]van-Ginneken B, Schaefer-Prokop C, Prokop M: Computer-aided diagnosis: how to move from the laboratory to the clinic. Radiology 2011, 261:719-32.
• [12]Jing Z, Bin L, Lianfang T. Lung nodule classification combining rule-based and SVM. In: Edited by Li K, Proceedings of the IEEE Fifth International Conference on Bio-Inspired Computing: Theories and Applications: 23–26 September 2010; Changsha, China. Piscataway, NJ: IEEE Computer Society; 2010. p. 1033–36
• [13]Kumar SA, Ramesh J, Vanathi PT, Gunavathi K. Robust and automated lung nodule diagnosis from CT images based on fuzzy systems. In: Edited by Manikandan V, Proceedings of the IEEE International Conference on Process Automation, Control and Computing: 20–22 July 2011; Coimbatore, India. Piscataway, NJ: IEEE Women in Engineering; 2011. p. 1–6
• [14]Chen H, Zhang J, Xu Y, Chen B, Zhang K: Performance comparison of artificial neural network and logistic regression model for differentiating lung nodules on CT scans. Expert Syst Appl 2012, 39:11503-9.
• [15]Kumar A, Mukhopadhyay S, Khandelwal N. 3d texture analysis of solitary pulmonary nodules using co-occurrence matrix from volumetric lung CT images. In: Edited by Sidky E, Proceedings of the SPIE Medical Imaging Conference: 28 February 2013; Lake Buena Vista, Florida, USA. Bellingham, WA: SPIE; 2013. p. 1–4
• [16]Keshani M, Azimifar Z, Tajeripour F, Boostani R: Lung nodule segmentation and recognition using SVM classifier and active contour modeling: a complete intelligent system. Comput Biol Med 2013, 43:287-300.
• [17]Lee M, Boroczky L, Sungur K, Cann A, Borczuk A, Kawut S, et al.: Computer-aided diagnosis of pulmonary nodules using a two-step approach for feature selection and classifier ensemble construction. Artif Intell Med 2010, 50:43-53.
• [18]Anand, SKV. Segmentation coupled textural feature classification for lung tumor prediction. In: Edited by Moses C, Proceedings of the IEEE International Conference on Communication Control and Computing Technologies: 7–9 October 2010; Tamil Nadu, India. IEEE Computer Society; 2010. p. 518–24.
• [19]Dmitriy Z, Feigenbaum J, Jacob F, Raicu D. Probabilistic lung nodule classification with belief decision trees. In: Edited by Liang Z, Proceedings of the 33rd Annual International Conference of the IEEE Engineering in Medicine and Biology Society: 30 August-3 September 2011; Boston, USA. IEEE; 2011. p. 4493–98.
• [20]Zhang F, Song Y, Cai W, Lee M, Zhou Y, Huang H, et al.: Lung nodule classification with multilevel patch-based context analysis. IEEE T Bio-Med Eng 2014, 61:1155-66.
• [21]Kuruvilla J, Gunavathi K: Lung cancer classification using neural networks for CT images. Comput Meth Prog Bio 2014, 113:202-9.
• [22]ELCAP Public lung image database. [http://www.via.cornell.edu/databases/lungdb.html]
• [23]Armato S, McLennan G, Bidaut L, McNitt-Gray M, Meyer C: The lung image database consortium (LIDC) and image database resource initiative (IDRI): a completed reference database of lung nodules on CT scans. Med Phys 2011, 38:915-31.
• [24]Farag A, Graham J, Farag A. Deformable models for random small-size objects: case of lung nodules in CT tomography. In: Edited by Ward R, Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing: 26–31 May 2013; Canada. Piscataway, NJ: IEEE Signal Processing Society; 2014. p. 1090–93.
• [25]Xie Y, Padgett J, Biancardi A, Reeves A: Automated aorta segmentation in low-dose chest CT images. Int J Comput Ass Rad 2014, 9:211-9.
• [26]Van de Wouwer G, Scheunders P, Van Dyck D: Statistical texture characterization from discrete wavelet representations. IEEE T Image Process 1999, 8:592-8.
• [27]Banerji S, Sinha A, Liu C: New image descriptors based on color, texture, shape, and wavelets for object and scene image classification. Neurocomput 2013, 117:173-85.
• [28]Zheng Z, Washington S: On selecting an optimal wavelet for detecting singularities in traffic and vehicular data. Transport Res C-Emer 2012, 25:18-33.
• [29]Singh R, Khare A: Fusion of multimodal medical images using daubechies complex wavelet transform: a multiresolution approach. Inform Fusion 2014, 19:49-60.
• [30]Al-Kadi O: Assessment of texture measures susceptibility to noise in conventional and contrast enhanced computed tomography lung tumour images. Comput Med Imag Grap 2010, 34:494-503.
• [31]Zhu Y, Tan Y, Hua Y, Wang M, Zhang G, Zhang J: Feature selection and performance evaluation of support vector machine (svm)-based classifier for differentiating benign and malignant pulmonary nodules by computed tomography. J Digit Imaging 2010, 23:51-65.
• [32]Wang J, Wu H, Sun T, Li X, Wang W, Tao L, et al.: Prediction models for solitary pulmonary nodules based on curvelet textural features and clinical parameters. Asian Pac J Cancer P 2013, 14:6019-23.
• [33]Wu H, Sun T, Wang J, Li X, Wang W, Huo D, et al.: Combination of radiological and gray level co-occurrence matrix textural features used to distinguish solitary pulmonary nodules by computed tomography. J Digit Imaging 2013, 26:797-802.
• [34]Sun T, Wang J, Li X, Lv P, Liu F, Luo Y, et al.: Comparative evaluation of support vector machines for computer aided diagnosis of lung cancer in CT based on a multi-dimensional data set. Comput Meth Prog Bio 2013, 111:519-24.
• [35]Raghavendra S, Chandra P: Support vector machine applications in the field of hydrology: a review. Appl Soft Comput 2014, 19:372-86.
• [36]Low F, Michel U, Dech S, Conrad C: Impact of feature selection on the accuracy and spatial uncertainty of per-field crop classification using support vector machines. ISPRS J Photogramm 2013, 85:102-19.
• [37]Valpine P, Bitter H, Brown M, Heller J: A simulation-approximation approach to simple size planning for high-dimensional classification studies. Biostatistics 2009, 10:424-35.
• [38]Boroczky L, Zhao L, Lee K: Feature subset selection for improving the performance of false positive reduction in lung nodule CAD. IEEE T Inf Technol B 2006, 10:504-11.
• [39]Gigliarano C, Figini S, Muliere P: Making classifier performance comparisons when ROC curves intersect. Comput Stat Data An 2014, 77:300-12.
• [40]Guvenir H, Kurtcephe M: Ranking instances by maximizing the area under ROC curve. IEEE T Knowl Data En 2012, 25:2356-66.