Masses In Mammograms Research Articles

Classification of Masses in Digital Mammograms Using the Genetic Ensemble Method

Abstract All over the world, breast cancer is the second leading cause of death in women above 40 years of age. To design an efficient classification system for breast cancer diagnosis, one has to use efficient algorithms for feature selection to reduce the feature space of mammogram classification. The current work investigates the use of hybrid genetic ensemble method for feature selection and classification of masses. Genetic algorithm (GA) is used to select a subset of features and to evaluate the fitness of the selected features, Adaptive boosting (AdaBoost) and Random Forest (RF) ensembles with 10-fold cross-validation are employed. The selected features are used to classify masses into benign or malignant using AdaBoost, RF, and single Decision Tree (DT) classifiers. The performance evaluation of classifiers indicates that AdaBoost outperforms both RF and single DT classifiers. AdaBoost achieves an accuracy of 96.15%, with 97.32% sensitivity, 95.90% specificity, and area under curve of AZ = 0.982 ± 0.004. The results obtained with the proposed method are better when compared with extant research work.

Computer-aided diagnosis of mammographic masses using local geometric constraint image retrieval

Computer-Aided Diagnosis of masses in mammograms is an important indicator of breast cancer. The use of retrieval systems in breast examination is gradually increasing. Hence, in mammographic mass retrieval, the method of exploiting the vocabulary tree framework and the inverted file has been proven to have high accuracy and excellent scalability. However, it only considered the features in each image as a visual word and ignored the spatial configurations of features, which greatly affects the retrieval performance. To overcome this drawback, we adopt the geometric verification method in mammographic mass retrieval. First, we obtain corresponding match features based on the vocabulary tree framework and the inverted file. Then, we grasp local similarity characteristics of deformations within the local regions by constructing the circle regions of corresponding pairs. Meanwhile, we quarter the circle to express the geometric relationship of local matches in the area and strictly generate the spatial encoding. Additionally, we rotate local matches in the area to control the strictness of geometric constraints. Finally, we verify geometric consistency to filter the false matches.The experimental results demonstrate that our method could significantly improve the retrieval accuracy with low computational cost.

Breast cancer detection based on Gabor-wavelet transform and machine learning methods

Among the major causes of female mortality, breast cancer used to pose big challenges to the medical world. Currently, the most popular method of monitoring and diagnoses—in addition to mammography—is carrying out repeated biopsies to locate the tumor further, that may result in loss of breast tissues. This paper presents an effective method of classifying and detecting the masses in mammograms. In the proposed method, we first attain the feature vector pertaining to each mammography image based on Gabor wavelet transform. Then, we performed tenfold cross validation through several experiments, analyzing the data complexity on each fold. We also used some machine learning methods as decision-making stage and achieved mean accuracies above 0.939, mean sensitivities as high as 0.951, and the mean specificities greater than 0.92. Evaluations and comparisons witness the effectiveness of the proposed method for better diagnosis of breast cancer against the known classification techniques developed in mammography. Simplicity, robustness and high accuracy are advantages of the proposed method.

Breast Mass Segmentation Using a Semi-automatic Procedure Based on Fuzzy C-means Clustering

Mammography is the primary modality that helped in the early detection and diagnosis of women breast diseases. Further, the process of extracting the masses in mammogram represents a challenging task facing the radiologists, due to problems such as fuzzy or speculated borders, low contrast and the presence of intensity inhomogeneities. Aims to help the radiologists in the diagnosis of breast cancer, many approaches have been conducted to automatically segment the masses in mammograms. Towards this aim, in this paper, we present a new approach for extraction of tumors from region-of-interest (ROI) using the algorithm of Fuzzy C-Means (FCM) setting two clusters for semi-automated segmentation. The proposed method meant to select as input data the set of pixels that enable to get the meaningful information required to segment the masses with high accuracy. This could be accomplished through eliminating unnecessary pixels, which influence on this process through separating it outside of the input data using an optimal threshold given by monitoring the change of clusters rate during the process of threshold decrementing. The proposed methodology has successfully segmented the masses, with an average sensitivity of 82.02% and specificity of 98.23%.

Method of differentiation of benign and malignant masses in digital mammograms using texture analysis based on phylogenetic diversity

Breast cancer is a disease resulting from the multiplication of abnormal breast cells, which form masses. Every year, breast cancer kills more than 500,000 women around the world. In 2015, 570,000 women died of breast cancer. When detected early, the five-year survival rate for breast cancer exceeds 80% of cases. Early diagnosis of breast cancer is critical for the survival of the patient. Screening by mammography is the most promising means for early diagnosis. This article presents a method of classifying malignant and benign breast tissue using digital mammography exams. This method employs texture descriptors from all image regions, including to the inner regions. This approach enables a more detailed texture description of the analyzed region of interest. The feature extraction is based on phylogenetic indexes. Then, classification is conducted using multiple classifiers. Experiments are performed to verify the performance of the proposed method. Results show that the method achieves 99.73% accuracy, 99.41% sensitivity, 99.84% specificity, and a receiver operating characteristic (ROC) curve with a value of one when using images of the Digital Database for Screening Mammography. An accuracy of 100% is achieved when using the Mammography Imaging Analysis Society image database. The use of phylogenetic indexes to describe patterns in regions of mammography images in both external and internal areas is thus effective in the categorization of malignant and benign tumors, thereby making the proposed method a robust tool for specialists.

Wavelet-like selective representations of multidirectional structures: a mammography case

The subject of this paper is selective representation of informative texture directionality in sparse domain of four verified multiscale transforms: contourlets, curvelets, tensor and complex wavelets. Directionality of linear or piecewise linear structures is a fundamental property in recognition of anatomical structures, i.e. separating the brain regions, directional characteristics of small coronary arteries. Another important example is spicule extraction in mammograms which is based on directional analysis of malignant spiculated lesions. The originality of the proposed experiments lies in optimization of multiscale wavelet-like representations to differentiate multidirectional structures of architectural distortions and spiculated masses in low-contrast noisy mammograms. For that purpose, the applied method consists in proposed phantom-based normalization and defined assessment criteria of directional information activity. The directional activity including both measures of angular resolution and angular selectivity was determined relative to increased background density and noise simulating reduced perceptibility of the analyzed objects. A numerically modeled phantom of multidirectional linear structures was used to assess subjectively and objectively the efficiency of nonlinear approximation for radially diverging spicule-like structures. Size and distribution of the structures simulates essential nature of spiculated signs of breast cancer. Basing on the experimental results, the complex wavelet domain was concluded to be the most effective tool to uniquely represent relevant information in the form of multidirectional piecewise linear structures in low-contrast noisy mammograms.

SCM-motivated enhanced CV model for mass segmentation from coarse-to-fine in digital mammography

A novel approach for mass segmentation from coarse-to-fine in digital mammography, termed as SCM-motivated enhanced CV algorithm, is presented in this paper. As well known, it is difficult to robustly achieve mammogram mass segmentation due to low contrast between normal and lesion tissues, as well as high density tissue interference in mammograms. Therefore, Spiking Cortical Model with biology background is introduced to achieve mammary-specific and mass edge detection, and this mass candidate is regarded as the initial contour of improved CV model followed by, effectively overcoming the drawback that CV method is sensitive to the initial contour; especially, the enhanced CV model innovatively combines the techniques of physical imaging principle, and local region-scalable force, harvesting the coarse-to-fine mass boundary accurately. The proposed method is tested totally on 400 mammograms from two well-known digitized datasets (digital database for screening mammography and mammography image analysis society database), achieving the average detection rate of 93.25%. By comparing with the region-based model with bias field (Method 1) and typical CV model (Method 2), we can reach the conclusion that proposed method is outperform other methods, yielding the average sensitivity of 95.83%, specificity of 99.13%, dice similarity co-efficient of 92.21% and AUC of 98.02%. In addition, this method is verified on the mammograms from Gansu Provincial Cancer Hospital, the detection results reveal that our method can accurately detect the abnormal in clinical application.

Classification of Masses in Digital Mammograms Using Firefly based Optimization

Classification of Masses in Digital Mammograms Using Firefly based Optimization

Morphological operation and scaled Réyni entropy based approach for masses detection in mammograms

Detection of suspicious masses in mammograms play a vital role in early diagnosis of breast cancer, to reduce the death rate among women. The presence of masses and calcification’s in mammograms is distinguishing signs for breast cancer diagnosis. But in some cases, due to contrast variation, fuzzy boundaries and presence of noise in mammograms, segmentation and detection of masses is challenging assignment. This paper presents a new segmentation approach to detect masses in mammographic images. The proposed approach consist of artifacts elimination and pectoral region extraction, suspicious mass enhancement using dual morphological operation technique and finally, extraction of Regions of Interest (ROIs) from background using scaled Reyni entropy. The proposed system has been tested on two data-sets i.e. mini- Mammographic Image Analysis Society (mini-MIAS) and Digital Database for Screening Mammography (DDSM), over 50 and 90 mammograms respectively. Performance achieved with proposed system in terms of True Positive Fraction (TPF) yields 93.2% and 93.9% respectively, at the rate of 1.48 and 0.74 average False Positive per Image (FP/I), tested on both Cranio-Caudal (CC) and Medio-Lateral Oblique (MLO) views. The obtained experimental results demonstrates that proposed method gives improved results for mass detection and can be useful for radiologists in diagnosis of breast cancer at early stage.

Classification of masses in digital mammograms using Biogeography-based optimization technique

Breast cancer is a leading cause of death in women in both developed and developing countries. Design and development of computer-based systems can assist radiologists in the effective treatment of breast cancer. For the design of an efficient classification system, efficient feature selection techniques must be used to reduce complexity of feature space in digital mammogram classification. The proposed methodology aims to explore use of Biogeography-based optimization to select a subset of features. Adaptive neuro-fuzzy inference system and artificial neural network are employed to evaluate fitness of the selected features. The features selected are used to train and test adaptive neuro-fuzzy inference system and artificial neural network classifiers. The experiment employed over 651 mammograms. The classification results shows that Biogeography-based optimization with adaptive neuro-fuzzy inference system is superior to Biogeography-based optimization with artificial neural network. Adaptive neuro-fuzzy inference system classifier achieve an accuracy of 98.92% with sensitivity of 99.10%, specificity of 98.72% and area under curve AZ = 0.999 ± 0.000. Outcomes achieved with the proposed Biogeography-based optimization with adaptive neuro-fuzzy inference system are far better as compared to some recent work.

Simultaneous detection and classification of breast masses in digital mammograms via a deep learning YOLO-based CAD system

Background and objectiveAutomatic detection and classification of the masses in mammograms are still a big challenge and play a crucial role to assist radiologists for accurate diagnosis. In this paper, we propose a novel Computer-Aided Diagnosis (CAD) system based on one of the regional deep learning techniques, a ROI-based Convolutional Neural Network (CNN) which is called You Only Look Once (YOLO). Although most previous studies only deal with classification of masses, our proposed YOLO-based CAD system can handle detection and classification simultaneously in one framework. MethodsThe proposed CAD system contains four main stages: preprocessing of mammograms, feature extraction utilizing deep convolutional networks, mass detection with confidence, and finally mass classification using Fully Connected Neural Networks (FC-NNs). In this study, we utilized original 600 mammograms from Digital Database for Screening Mammography (DDSM) and their augmented mammograms of 2,400 with the information of the masses and their types in training and testing our CAD. The trained YOLO-based CAD system detects the masses and then classifies their types into benign or malignant. ResultsOur results with five-fold cross validation tests show that the proposed CAD system detects the mass location with an overall accuracy of 99.7%. The system also distinguishes between benign and malignant lesions with an overall accuracy of 97%. ConclusionsOur proposed system even works on some challenging breast cancer cases where the masses exist over the pectoral muscles or dense regions.

Subclass based parallel learning neural network for classification of masses in mammograms

Computer aided detection assists radiologists by providing second opinion in the mammography detection, and reduce misdiagnosis. An expert system with novel subclass based learning multiple neural network classifier (SBLMNN) has been proposed to solve the mammogram mass classification problem. This work explores the significance of the modular learning in artificial neural networks, inspired from the visual cortex basis of human learning. It is a two stage learning process. In stage I, the proposed architecture processes parallel on the radiological characteristics of mass like shape, margin and texture features in separate modules similar to the visual cortex to identify the subclasses. The intermediate outputs of the independent modules are processed to classify the mass into benign or malignant in stage II. Modularization and deep learning considered in the proposed method improves the performance of the classifier and speed of learning. For the experimental analysis, images are obtained from the mammogram image analysis society. The experiments were implemented in MATLAB. For benign and malignant classification, the shows that SBLMNN accuracy is 92%, which is higher than monolithic MLP neural network architecture.

Computer aided breast cancer diagnosis: an SVM-based mammogram classification approach

This paper addresses the problem of mammogram classification (benign vs. malignant) through the utilisation of the state-of-the-art machine learning paradigm of support vector machines (SVMs). We are particularly interested in evaluating the discrimination efficiency of a set of feature extraction algorithms that have been proposed in the relevant literature for describing the textual characteristics present in mammogram masses. Our research focuses on comparing the classification accuracy associated with two well-established feature generation methodologies, namely, spatial grey level dependence method (SGLDM) and run difference method (RDM). Our experimentation was conducted on a publicly available mammogram database by parameterising the underlying kernel function of SVMs on different subsets of features. Our results indicate a moderately high classification accuracy for the linear SVM classifier when trained on the complete set of features.

Computer aided breast cancer diagnosis: an SVM-based mammogram classification approach

This paper addresses the problem of mammogram classification (benign vs. malignant) through the utilisation of the state-of-the-art machine learning paradigm of support vector machines (SVMs). We are particularly interested in evaluating the discrimination efficiency of a set of feature extraction algorithms that have been proposed in the relevant literature for describing the textual characteristics present in mammogram masses. Our research focuses on comparing the classification accuracy associated with two well-established feature generation methodologies, namely, spatial grey level dependence method (SGLDM) and run difference method (RDM). Our experimentation was conducted on a publicly available mammogram database by parameterising the underlying kernel function of SVMs on different subsets of features. Our results indicate a moderately high classification accuracy for the linear SVM classifier when trained on the complete set of features.

Classification of Breast Masses in Mammograms Using 2D Homomorphic Transform Features and Supervised Classifiers

Classification of Breast Masses in Mammograms Using 2D Homomorphic Transform Features and Supervised Classifiers

Hybrid methods for feature extraction for breast masses classification

This paper is focusing on feature extraction methods for malignant masses in mammograms and its classification. It proposes seven texture features for GLCM method and to be applied on sub-images to enhance its performance. It also proposes three hybrid methods named Wavelet-CT1, Wavelet-CT2 and ST-GLCM. The three hybrid methods are merging two types of different features. In this research, we divide the region of interest image into s×s sub-images and a contrast stretching stage is applied before extracting the features from each sub-image. This research also introduces two Contourlet methods (CT1 and CT2). The feature extraction methods are applied on each sub-image of ROI. CT1 is applying Contourlet at level 4. CT2 is applying Contourlet at levels [4321]. GLCM uses seven texture features. Wavelet-CT1 is applying CT1 method to all bands of wavelet coefficients at level one. Wavelet-CT2 is merging high frequency bands of wavelet at level one with contourlet coefficients of CT2. ST-GLCM merges seven statistical features and seven texture features extracted from Grey level Co-occurrence Matrix (GLCM). The proposed methods are compared with multi-resolution feature extraction methods using discrete wavelet, ridgelet and curvelet transform. SVM is used for classification. Images from Digital Database for Screening Mammography (DDSM) and Mammograms Image Analysis Society (MIAS) database are used for evaluation. The performance of proposed methods ST-GLCM, GLCM, Wavelet-CT1 and Contourlet (CT2) outperform all current existing feature extraction methods in terms of AUC measure. The extracted number of features by using GLCM or ST-GLCM is small compared to multi-resolution features.

Detection and classification of the breast abnormalities in digital mammograms via regional Convolutional Neural Network.

Automatic detection and classification of the masses in mammograms are still a big challenge and play a crucial role to assist radiologists for accurate diagnosis. In this paper, we propose a novel computer-aided diagnose (CAD) system based on one of the regional deep learning techniques: a ROI-based Convolutional Neural Network (CNN) which is called You Only Look Once (YOLO). Our proposed YOLO-based CAD system contains four main stages: mammograms preprocessing, feature extraction utilizing multi convolutional deep layers, mass detection with confidence model, and finally mass classification using fully connected neural network (FC-NN). A set of training mammograms with the information of ROI masses and their types are used to train YOLO. The trained YOLO-based CAD system detects the masses and classifies their types into benign or malignant. Our results show that the proposed YOLO-based CAD system detects the mass location with an overall accuracy of 96.33%. The system also distinguishes between benign and malignant lesions with an overall accuracy of 85.52%. Our proposed system seems to be feasible as a CAD system capable of detection and classification at the same time. It also overcomes some challenging breast cancer cases such as the mass existing in the pectoral muscles or dense regions.

Neighborhood Structural Similarity Mapping for the Classification of Masses in Mammograms.

In this paper, two novel feature extraction methods, using neighborhood structural similarity (NSS), are proposed for the characterization of mammographic masses as benign or malignant. Since gray-level distribution of pixels is different in benign and malignant masses, more regular and homogeneous patterns are visible in benign masses compared to malignant masses; the proposed method exploits the similarity between neighboring regions of masses by designing two new features, namely, NSS-I and NSS-II, which capture global similarity at different scales. Complementary to these global features, uniform local binary patterns are computed to enhance the classification efficiency by combining with the proposed features. The performance of the features are evaluated using the images from the mini-mammographic image analysis society (mini-MIAS) and digital database for screening mammography (DDSM) databases, where a tenfold cross-validation technique is incorporated with Fisher linear discriminant analysis, after selecting the optimal set of features using stepwise logistic regression method. The best area under the receiver operating characteristic curve of 0.98 with an accuracy of is achieved with the mini-MIAS database, while the same for the DDSM database is 0.93 with accuracy .

Bi-model processing for early detection of breast tumor in CAD system

Early screening of skeptical masses in mammograms may reduce mortality rate among women. This rate can be further reduced upon developing the computer-aided diagnosis system with decrease in false assumptions in medical informatics. This method highlights the early tumor detection in digitized mammograms. For improving the performance of this system, a novel bi-model processing algorithm is introduced. It divides the region of interest into two parts, the first one is called pre-segmented region (breast parenchyma) and other is the post-segmented region (suspicious region). This system follows the scheme of the preprocessing technique of contrast enhancement that can be utilized to segment and extract the desired feature of the given mammogram. In the next phase, a hybrid feature block is presented to show the effective performance of computer-aided diagnosis. In order to assess the effectiveness of the proposed method, a database provided by the society of mammographic images is tested. Our experimental outcomes on this database exhibit the usefulness and robustness of the proposed method.

Globally supported radial basis function based collocation method for evolution of level set in mass segmentation using mammograms

Computer-aided detection systems play an important role for the detection of breast abnormalities using mammograms. Global segmentation of mass in mammograms is a complex process due to low contrast mammogram images, irregular shape of mass, speculated margins, and the presence of intensity variations of pixels. This work presents a new approach for mass detection in mammograms, which is based on the variational level set function. Mesh-free based radial basis function (RBF) collocation approach is employed for the evolution of level set function for segmentation of breast as well as suspicious mass region. The mesh-based finite difference method (FDM) is used in literature for evolution of level set function. This work also showcases a comparative study of mesh-free and mesh-based approaches. An anisotropic diffusion filter is employed for enhancement of mammograms. The performance of mass segmentation is analyzed by computing statistical measures. Binarized statistical image features (BSIF) and variants of local binary pattern (LBP) are computed from the segmented suspicious mass regions. These features are given as input to the supervised support vector machine (SVM) classifier to classify suspicious mass region as mass (abnormal) or non-mass (normal) region. Validation of the proposed algorithm is done on sample mammograms taken from publicly available Mini-mammographic image analysis society (MIAS) and Digital Database for Screening Mammography (DDSM) datasets. Combined BSIF features perform better as compared to LBP variants with the performance reported as 97.12% sensitivity, 92.43% specificity, and 98% AUC with 5.12 FP/I on DDSM dataset; and 95.12% sensitivity, 92.41% specificity, and 95% AUC with 4.01FP/I on MIAS dataset.