Search

Determination of breast tumors from B-Mode Ultrasound (US) image is a perplexing one. Researches employing statistical modeling such as Nakagami, Normal Inverse Gaussian (NIG) distributed parametric images in this classification task have already explored but experimentation of those statistical models on contourlet transformed coefficient image in breast tumor classification task has not reported yet. The proposed method is established by considering 250 clinical cases from a publicly available database. In this database each clinical case exists as *.bmp format. In the preprocessing step firstly, the ultrasound B-Mode image is binarized to detect the lesion contour. Then contourlet transformation is employed. These contourlet sub band coefficients are shown to be modeled effectively by Nakagami and NIG distributions. These Nakagami and NIG parametric images are obtained by estimating the parameters of those prior statistical distributions locally. Few shape and statistical features are chosen according to their effectiveness on those parametric images. The benign and malignant breast tumors are classified utilizing these features with different classifiers such as the support vector machine, k-nearest neighbors, fitted binary classification decision tree, binary Gaussian kernel classification model, linear classification models for binary learning with high-dimensional etc. It is observed that classification performance of NIG statistical model based parametric version of contourlet coefficient images gained better accuracy than those of Nakagami statistical model. GUB JOURNAL OF SCIENCE AND ENGINEERING, Vol 7, Dec 2020 P 21-26

Automated visual identification of Benign and Malignant breast tumors even with ultrasound (US) B-Mode image is still an open area of research. This paper presents parametric image-based approach to classify and detect benign and malignant breast tumors from ultrasound images using a custom-made convolutional neural network architecture. The Rician Inverse Gaussian (RiIG) distribution is presented here as a suitable model for describing the statistics of the ultrasound images in the Contourlet Transform domain. Locally computed values of the dispersion parameters of the RiIG distribution in various contourlet sub-bands yield parametric images those are classified using the proposed convolutional neural network. Experiments are conducted on a publicly available dataset of 250 images, of 100 belong to the benign Fibroadenoma and 150 in the malignant category. The proposed method provides accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive values (NPV) of 96%, 97.3%, 94.12%, 96% and 96%, respectively. It is also shown that the accuracy obtained by the Proposed Method is higher than several recently reported results.

Classification of Benign and Malignant breast cancer tumours by B-Mode ultrasound image analysis is a challenging problem. In this study, first suitable raw radio frequency (RF) frame is selected based on strain and velocity imaging. The consequent B-Mode ultrasound (US) image is binarized and subsequently subjected to contourlet transform. In the proposed method only six contourlet sub-band coefficients are considered from the binary image to reduce the computation time duration and then texture and statistical features are extracted from the contourlet coefficient image. Finally, the lesion is classified as benign or malignant using three classifiers such as artificial neural network (ANN), K-Nearest Neighbours (KNN) and Support Vector Machine (SVM). It is shown that the proposed method yields classification accuracies of 95%, 92.5% and 90% over a competent computational time by using ANN, KNN and SVM classifiers respectively.

Speckle noise is an inherent phenomenon in medical ultrasound (US) images. Since it degrades an ultrasound image quality and reduces its diagnostic value, reduction of speckle noise is a very important pre-processing step in ultrasound image processing. For this purpose, the knowledge of the statistics of speckle noise is necessary; especially in the multi-resolution transform domain due to their sparse and efficient representation of images. In this paper a Bessel K-Form (BKF) probability density function (pdf) is proposed as a highly suitable prior for modeling the log-transformed speckle noise in the well-known contourlet transform domain. A maximum likelihood based method is presented for estimating the parameters of the BKF pdf. The appropriateness of the BKF pdf in modeling the speckle is studied for different noise levels in the contourlet transform domain, in addition the suitability of BKF model is investigated for the case of real US images. It is shown that, in general the BKF can model the statistics of the contourlet transform coefficients corresponding to log-transformed speckle better than the traditional Gaussian and normal inverse Gaussian pdfs.

Speckle noise in medical ultrasound (US) degrades the image quality and reduces its diagnostic value. Reduction of speckle noise is an important pre-processing step for the analysis and processing of medical ultrasound images. Knowledge of the statistics of the log-transformed speckle especially in the multi-resolution transform domain is important for developing effective homomorphic despeckling techniques, the most popular approach of speckle reduction from ultrasound images. In this paper, the bessel K-form (BKF) probability density function (pdf) is proposed as a highly suitable prior for modeling the log-transformed speckle noise in the well-known contourlet transform domain. A maximum likelihood based method is introduced for estimating the parameters of the BKF pdf. The effectiveness of the proposed estimation method is demonstrated using Monte Carlo simulations. The appropriateness of the BKF pdf in modeling the speckle is first studied extensively for simulated noise of different levels in the contourlet transform domain. Next, the suitability of BKF model is investigated for the case of real US images that include neonatal brain and breast tumors. It is shown that, in general the BKF prior can model the statistics of the contourlet transform coefficients corresponding to the log-transformed speckle better than the traditionally used Gaussian, normal inverse Gaussian and generalized Nakagami pdfs.