Application of neural networks for the analysis of intravascular ultrasound and histological aortic wall appearance-an in vitro tissue characterization study.

The role of tissue characterization by intravascular ultrasound (IVUS) imaging of the aortic wall has not been well established. The artificial neural networks (ANNs) are a promising tool for image classification. The aim of the study was to assess the texture correlation between matching IVUS and histologic images of the aortic wall. The computer-based discrimination of pathology within the data sets was also evaluated. In vitro IVUS images and histologic sections from 36 aortic segments were compared using texture parameters that produced the best correlation or the highest discriminative value. The images were classified as normal or abnormal with variable degrees of pathology. Tissue characterization was performed by a nearest neighbor classifier, linear discriminant analysis (LDA) and the ANN-based approach. Good agreement was observed between IVUS and the histologic reference with a correlation coefficient of r = 0.89, r = 0.76 and r = 0.71 for the three most successful texture parameters. The ANN-based approach was the most effective in discriminant analysis, with a correct classification rate of 87.5% for histologic images and 79.2% for IVUS data. The study shows that ANNs are a potentially effective tool for assessment of IVUS aortic images.

[1]  M Kaps,et al.  Variability of Doppler microembolic signal counts in patients with prosthetic cardiac valves. , 1995, Stroke.

[2]  Keinosuke Fukunaga,et al.  Introduction to Statistical Pattern Recognition , 1972 .

[3]  Rama Chellappa,et al.  Estimation and choice of neighbors in spatial-interaction models of images , 1983, IEEE Trans. Inf. Theory.

[4]  S C Horii,et al.  Improving the distinction between benign and malignant breast lesions: the value of sonographic texture analysis. , 1993, Ultrasonic imaging.

[5]  R. Linder,et al.  Neural network approach for computer-assisted interpretation of ultrasound images of the gallbladder. , 1993, European journal of radiology.

[6]  S. Momjian,et al.  Stratified Gray-Scale Median Analysis and Color Mapping of the Carotid Plaque: Correlation With Endarterectomy Specimen Histology of 28 Patients , 2005, Stroke.

[7]  Cedric Nishan Canagarajah,et al.  A robust automatic clustering scheme for image segmentation using wavelets , 1996, IEEE Trans. Image Process..

[8]  R. Kloner,et al.  Characterization of collagen by high-frequency ultrasound: evidence for different acoustic properties based on collagen fiber morphologic characteristics. , 1997, American heart journal.

[9]  S K Holland,et al.  Ultrasound tissue characterization of breast biopsy specimens. , 1991, Ultrasonic imaging.

[10]  J M Thijssen,et al.  Application of Neural Networks for the Classification of Diffuse Liver Disease by Quantitative Echography , 1993, Ultrasonic imaging.

[11]  J. Thomas,et al.  Quantification of intramural calcification in coronary intravascular ultrasound images with automated image analysis. , 1998, American heart journal.

[12]  Anil K. Jain,et al.  Artificial neural networks for feature extraction and multivariate data projection , 1995, IEEE Trans. Neural Networks.

[13]  Walter L. Henry,et al.  Intravascular Ultrasound Imaging of Human Coronary Arteries In Vivo: Analysis of Tissue Characterizations With Comparison to In Vitro Histological Specimens , 1991, Circulation.

[14]  C. Tracy,et al.  American College of Cardiology Clinical Expert Consensus Document on Standards for Acquisition, Measurement and Reporting of Intravascular Ultrasound Studies (IVUS). A report of the American College of Cardiology Task Force on Clinical Expert Consensus Documents. , 2001, Journal of the American College of Cardiology.

[15]  T C Gerber,et al.  Differentiation of Intracardiac Tumors and Thrombi by Echocardiography Tissue Characterization: Comparison of an Artificial Neural Network and Human Observers , 2000, Echocardiography.

[16]  S. Rahimtoola,et al.  Value of Digital Image Processing of Two-Dimensional Echocardiograms in Differentiating Active From Chronic Vegetations of Infective Endocarditis , 1988, Circulation.

[17]  S Swarnamani,et al.  Application of artificial neural networks for the classification of liver lesions by image texture parameters. , 1996, Ultrasound in medicine & biology.

[18]  S. M. Collins,et al.  Ultrasound characterization of acoustic properties of acute intracardiac thrombi: studies in a new experimental model. , 1988, Journal of the American Society of Echocardiography : official publication of the American Society of Echocardiography.

[19]  Andrzej Materka,et al.  Texture analysis for magnetic resonance imaging , 2006 .

[20]  S. Jeffrey,et al.  Ultrasound Tissue Characterization of Breast Biopsy Specimens: Expanded Study , 1996 .

[21]  P Sherman,et al.  Classification of ultrasonic image texture by statistical discriminant analysis of neutral networks. , 1991, Computerized medical imaging and graphics : the official journal of the Computerized Medical Imaging Society.

[22]  Robert M. Haralick,et al.  Textural Features for Image Classification , 1973, IEEE Trans. Syst. Man Cybern..

[23]  J.T.M. Verhoeven,et al.  Regular ArticleApplication of Neural Networks for the Classification of Diffuse Liver Disease by Quantitative Echography , 1993 .

[24]  P. Yock,et al.  Intravascular ultrasound: novel pathophysiological insights and current clinical applications. , 2001, Circulation.

[25]  Michael C. Fishbein,et al.  Histopathologic Validation of Angioscopy and Intravascular Ultrasound , 1991, Circulation.

[26]  Jürgen Schürmann,et al.  Pattern classification , 1996 .

[27]  R.M. Haralick,et al.  Statistical and structural approaches to texture , 1979, Proceedings of the IEEE.