Open Access Open Access  Restricted Access Subscription Access

Classification of Liver Cirrhosis with Statistical Analysis of Texture Parameters

Hafeez Ullah Janjua, Farah Andleeb, Sidra Aftab, Fayyaz Hussain, Ghulam Gilanie

Abstract


Attempts have been made to study texture parameters for differentiation of the images. This paper aims at the classification of normal and abnormal (i.e. cirrhosis) ultrasound liver images through custom developed feature extractor software (FES). To classify the slices as normal and abnormal, different regions of interest (ROI) are processed to extract texture features based on statistical moments. Experiments reveal that ROI of size 64 × 64 is best for liver images classification. Statistical texture features like standard deviation, kurtosis, skewness, flatness, entropy and energy are extracted via FES. There is a significant variation in values of these seven parameters for abnormal images as compared with healthy ones. A machine learning tool, Waikato Environment Knowledge Analysis (Weka), has been used to verify the standard evaluation parameters by calculating precision, mean error, kappa statistics, ROC area, TP rate and FP rate. This tool shows an excellent agreement with our derived results. Overall, the proposed method of classification of normal and abnormal liver slices obtained accuracy of 96%. This work supports in visual identification of images using computer aided diagnostic tool that helps radiologist and doctors to diagnosing medical images automatically absolute results instantly can be found.

Full Text:

PDF

References


K. Ito, D.G. Mitchell. Hepatic morphologic changes in cirrhosis: MR imaging findings, Abdominal Imaging. 2000; 25(5): 456–61p.

B. Xia, H. Jiang, H. Liu, D. Yi. A novel hepatocellular carcinoma image classification method based on voting ranking random forests, Comput Math Methods Med. 2016; 2016: 8p.

Mimi, X. Gao. Medical Imaging and Informatics: Second International Conference. MIMI 2007, Beijing, China, August 14–16, 2007: revised selected papers, Berlin; New York: Springer; 2008. Available from: http://www.myilibrary.com?id=185512.

H. Awaya, D.G. Mitchell, T. Kamishima, G. Holland, K. Ito, T. Matsumoto. Cirrhosis: modified caudate–right lobe ratio, Radiology. 2002; 224(3): 769–74p.

M. Farshid Babapour, T.-F. Ali Abbaspour, Z. Reza Aghaeizadeh, A. Shahram, C. Yen-Wei. A novel wavelet based multi-scale statistical shape model-analysis for the liver application: segmentation and classification, Curr Med Imaging Rev. 2010; 6(3): 145–55p.

S. Zhou, J. Wan, eds. A survey of algorithms for the analysis of diffused liver disease from B-mode ultrasound images, 9th International Conference on Electronic Measurement & Instruments. 2009, 16–19 Aug.

A.Z.S.A. Zaid, M.W. Fakhr, A.F.A. Mohamed, eds. Automatic diagnosis of liver diseases from ultrasound images, International Conference on Computer Engineering and Systems. 2006, 5–7 Nov.

N. P, H. G, B. S, K. K. Feature extraction of mammograms, Int J Bioinform Res. 2012; 4(1): 241p.

E.B. Harald Lutz, editor. 1. Diagnostic Imaging. 2. Ultrasonography. 3. Pediatrics - Instrumentation. 2nd Edn., WHO Library Cataloguing-in-Publication Data; 2011.

S. Manikandan, V. Rajamani. A mathematical approach for feature selection & image retrieval of ultra sound kidney image databases, Eur J Sci Res. 2008; 24(2): 163–71p.

An investigation of the textural characteristics associated with gray level cooccurrence matrix statistical parameters, IEEE Trans Geosci Remote Sens. 1995; 33(2): 293–304p.

N.R. Razman, W.M.H.W. Mahmud, N.A. Shaharuddin, eds. Filtering technique in ultrasound for kidney, liver and pancreas image using Matlab, IEEE Student Conference on Research and Development (SCOReD). IEEE, 2015.

L. Rokach. Taxonomy for characterizing ensemble methods in classification tasks: a review and annotated bibliography, Comput Stat Data Anal. 2009; 53(12): 4046–72p.

D.E. Gray. Doing Research in the Real World. 2013. Available from: https://nls.ldls.org.uk/welcome.html?ark:/81055/vdc_100025413802.0x000001.

I.H. Witten, E. Frank. Data Mining Practical Machine Learning Tools and Techniques. San Francisco, CA: Elsevier/Morgan Kaufmann; 2005.

F.-B. Enrique, R. Daniel, G. Marcos, F.-L. Carlos, E. Norberto, M. Cristian Robert, et al. A hybrid evolutionary system for automated artificial neural networks generation and simplification in biomedical applications, Curr Bioinform. 2015; 10(5): 672–91p.


Refbacks

  • There are currently no refbacks.