Detection of breast masses in mammograms by density slicing and texture flow-field analysis.

We propose a method for the detection of masses in mammographic images that employs Gaussian smoothing and sub-sampling operations as preprocessing steps. The mass portions are segmented by establishing intensity links from the central portions of masses into the surrounding areas. We introduce methods for analyzing oriented flow-like textural information in mammograms. Features based on flow orientation in adaptive ribbons of pixels across the margins of masses are proposed to classify the regions detected as true mass regions or false-positives (FPs). The methods yielded a mass versus normal tissue classification accuracy represented as an area (Az) of 0.87 under the receiver operating characteristics (ROCs) curve with a dataset of 56 images including 30 benign disease, 13 malignant disease, and 13 normal cases selected from the mini Mammographic Image Analysis Society database. A sensitivity of 81% was achieved at 2.2 FPs/image. Malignant tumor versus normal tissue classification resulted in a higher Az value of 0.9 under the ROC curve using only the 13 malignant and 13 normal cases with a sensitivity of 85% at 2.45 FPs/image. The mass detection algorithm could detect all the 13 malignant tumors successfully, but achieved a success rate of only 63% (19/30) in detecting the benign masses. The mass regions that were successfully segmented were further classified as benign or malignant disease by computing five texture features based on gray-level co-occurrence matrices (GCMs) and using the features in a logistic regression method. The features were computed using adaptive ribbons of pixels across the boundaries of the masses. Benign versus malignant classification using the GCM-based texture features resulted in Az = 0.79 with 19 benign and 13 malignant cases.     

Gradient and texture analysis for the classification of mammographic masses

Computer-aided classification of benign and malignant masses on mammograms is attempted in this study by computing gradient-based and texture-based features. Features computed based on gray-level co-occurrence matrices (GCMs) are used to evaluate the effectiveness of textural information possessed by mass regions in comparison with the textural information present in mass margins. A method involving polygonal modeling of boundaries is proposed for the extraction of a ribbon of pixels across mass margins. Two gradient-based features are developed to estimate the sharpness of mass boundaries in the ribbons of pixels extracted from their margins. A total of 54 images (28 benign and 26 malignant) containing 39 images from the Mammographic Image Analysis Society (MIAS) database and 15 images from a local database are analyzed. The best benign versus malignant classification of 82.1%, with an area (Az) of 0.85 under the receiver operating characteristics (ROC) curve, was obtained with the images from the MIAS database by using GCM-based texture features computed from mass margins. The classification method used is based on posterior probabilities computed from Mahalanobis distances. The corresponding accuracy using jack-knife classification was observed to be 74.4%, with Az = 0.67. Gradient-based features achieved Az = 0.6 on the MIAS database and Az = 0.76 on the combined database. The corresponding values obtained using jack-knife classification were observed to be 0.52 and 0.73 for the MIAS and combined databases, respectively.     

A study on several machine-learning methods for classification of malignant and benign clustered microcalcifications

In this paper, we investigate several state-of-the-art machine-learning methods for automated classification of clustered microcalcifications (MCs). The classifier is part of a computer-aided diagnosis (CADx) scheme that is aimed to assisting radiologists in making more accurate diagnoses of breast cancer on mammograms. The methods we considered were: support vector machine (SVM), kernel Fisher discriminant (KFD), relevance vector machine (RVM), and committee machines (ensemble averaging and AdaBoost), of which most have been developed recently in statistical learning theory. We formulated differentiation of malignant from benign MCs as a supervised learning problem, and applied these learning methods to develop the classification algorithm. As input, these methods used image features automatically extracted from clustered MCs. We tested these methods using a database of 697 clinical mammograms from 386 cases, which included a wide spectrum of difficult-to-classify cases. We analyzed the distribution of the cases in this database using the multidimensional scaling technique, which reveals that in the feature space the malignant cases are not trivially separable from the benign ones. We used receiver operating characteristic (ROC) analysis to evaluate and to compare classification performance by the different methods. In addition, we also investigated how to combine information from multiple-view mammograms of the same case so that the best decision can be made by a classifier. In our experiments, the kernel-based methods (i.e., SVM, KFD, and RVM) yielded the best performance (Az = 0.85, SVM), significantly outperforming a well-established, clinically-proven CADx approach that is based on neural network (Az = 0.80).     

A concentric morphology model for the detection of masses in mammography

We propose a technique for the automated detection of malignant masses in screening mammography. The technique is based on the presence of concentric layers surrounding a focal area with suspicious morphological characteristics and low relative incidence in the breast region. Mammographic locations with high concentration of concentric layers with progressively lower average intensity are considered suspicious deviations from normal parenchyma. The multiple concentric layers (MCLs) technique was trained and tested using the craniocaudal views of 270 mammographic cases with biopsy proven malignant masses from the digital database of screening mammography. One-half of the available cases were used for optimizing the parameters of the detection algorithm. The remaining cases were used for testing. During testing, malignant masses were detected with 92%, 88%, and 81% sensitivity at 5.4, 2.4, and 0.6 false positive marks per image. Testing on 82 normal screening mammograms showed a false positive rate of 5.0, 1.7, and 0.2 marks per image at the previously reported operating points. Furthermore, additional evaluation on 135 benign cases produced a significantly lower detection rate for benign masses (61.6%, 58.3%, and 43.7% at 5.1, 2.8, and 1.2 false positives per image, respectively). Overall, MCL is a promising computer-assisted detection strategy for screening mammograms to identify malignant masses while maintaining the detection rate of benign masses considerably lower.     

Optimizing Case-based detection performance in a multiview CAD system for mammography

When reading mammograms, radiologists combine information from multiple views to detect abnormalities. Most computer-aided detection (CAD) systems, however, use primitive methods for inclusion of multiview context or analyze each view independently. In previous research it was found that in mammography lesion-based detection performance of CAD systems can be improved when correspondences between MLO and CC views are taken into account. However, detection at case level detection did not improve. In this paper, we propose a new learning method for multiview CAD systems, which is aimed at optimizing case-based detection performance. The method builds on a single-view lesion detection system and a correspondence classifier. The latter provides class probabilities for the various types of region pairs and correspondence features. The correspondence classifier output is used to bias the selection of training patterns for a multiview CAD system. In this way training can be forced to focus on optimization of case-based detection performance. The method is applied to the problem of detecting malignant masses and architectural distortions. Experiments involve 454 mammograms consisting of four views with a malignant region visible in at least one of the views. To evaluate performance, five-fold cross validation and FROC analysis was performed. Bootstrapping was used for statistical analysis. A significant increase of case-based detection performance was found when the proposed method was used. Mean sensitivity increased by 4.7% in the range of 0.01-0.5 false positives per image.     

Improvement of sensitivity of breast cancer diagnosis with adaptiveneighborhood contrast enhancement of mammograms

Mammograms are difficult to interpret, especially of cancers at their early stages. We analyze the effectiveness of our adaptive neighborhood contrast enhancement (ANCE) technique in increasing the sensitivity of breast cancer diagnosis. Seventy-eight screen-film mammograms of 21 difficult cases (14 benign and seven malignant), 222 screen-film mammograms of 28 interval cancer patients and six benign control cases were digitized with a high-resolution of about 4096×2048×10-bit pixels and then processed with the ANCE method. Unprocessed and processed digitized mammograms as well as the original films were presented to six experienced radiologists for a receiver operating characteristic (ROC) evaluation for the difficult case set and to three reference radiologists for the interval cancer set. The results show that the radiologists' performance with the ANCE-processed images is the best among the three sets of images (original, digitized, and enhanced) in terms of area under the ROC curve and that diagnostic sensitivity is improved by the ANCE algorithm. All of the 19 interval cancer cases not detected with the original films of earlier mammographic examinations were diagnosed as malignant with the corresponding ANCE-processed versions, while only one of the six benign cases initially labeled correctly with the original mammograms was interpreted as malignant after enhancement. This study demonstrates the potential for improvement of diagnostic performance in early detection of breast cancer with digital image enhancement     

Temporal change analysis for characterization of mass lesions in mammography

In this paper, we present a fully automated computer-aided diagnosis (CAD) program to detect temporal changes in mammographic masses between two consecutive screening rounds. The goal of this work was to improve the characterization of mass lesions by adding information about the tumor behavior over time. Towards this goal we previously developed a regional registration technique that finds for each mass lesion on the current view a location on the prior view where the mass was most likely to develop. For the task of interval change analysis, we designed two kinds of temporal features: difference features and similarity features. Difference features indicate the (relative) change in feature values determined on prior and current views. These features may be especially useful for lesions that are visible on both views. Similarity features measure whether two regions are comparable in appearance and may be useful for lesions that are visible on the prior view as well as for newly developing lesions. We evaluated the classification performance with and without the use of temporal features on a dataset consisting of 465 temporal mammogram pairs, 238 benign, and 227 malignant. We used cross validation to partition the dataset into a training set and a test set. The training set was used to train a support vector machine classifier and the test set to evaluate the classifier. The average A(z) value (area under the receiver operating characteristic curve) for classifying each lesion was 0.74 without temporal features and 0.77 with the use of temporal features. The improvement obtained by adding temporal features was statistically significant (P = 0.005). In particular, similarity features contributed to this improvement. Furthermore, we found that the improvement was comparable for masses that were visible and for masses that were not visible on the prior view. These results show that the use of temporal features is an effective approach to improve the characterization of masses.     

Uncertainty in the output of artificial neural networks

Analysis of the performance of artificial neural networks (ANNs) is usually based on aggregate results on a population of cases. In this paper, we analyze ANN output corresponding to the individual case. We show variability in the outputs of multiple ANNs that are trained and "optimized" from a common set of training cases. We predict this variability from a theoretical standpoint on the basis that multiple ANNs can be optimized to achieve similar overall performance on a population of cases, but produce different outputs for the same individual case because the ANNs use different weights. We use simulations to show that the average standard deviation in the ANN output can be two orders of magnitude higher than the standard deviation in the ANN overall performance measured by the Az value. We further show this variability using an example in mammography where the ANNs are used to classify clustered microcalcifications as malignant or benign based on image features extracted from mammograms. This variability in the ANN output is generally not recognized because a trained individual ANN becomes a deterministic model. Recognition of this variability and the deterministic view of the ANN present a fundamental contradiction. The implication of this variability to the classification task warrants additional study.     

Polygonal modeling of contours of breast tumors with the preservation of spicules

Malignant breast tumors typically appear in mammograms with rough, spiculated, or microlobulated contours, whereas most benign masses have smooth, round, oval, or macrolobulated contours. Several studies have shown that shape factors that incorporate differences as above can provide high accuracies in distinguishing between malignant tumors and benign masses based upon their contours only. However, global measures of roughness, such as compactness, are less effective than specially designed features based upon spicularity and concavity. We propose a method to derive polygonal models of contours that preserve spicules and details of diagnostic importance. We show that an index of spiculation derived from the turning functions of the polygonal models obtained by the proposed method yields better classification accuracy than a similar measure derived using a previously published method. The methods were tested with a set of 111 contours of 65 benign masses and 46 malignant tumors. A high classification accuracy of 0.94 in terms of the area under the receiver operating characteristics curve was obtained.     

Computer-aided breast cancer detection and diagnosis of massesusing difference of Gaussians and derivative-based feature saliency

A new model-based vision (MBV) algorithm is developed to find regions of interest (ROI's) corresponding to masses in digitized mammograms and to classify the masses as malignant/benign. The MBV algorithm is comprised of 5 modules to structurally identify suspicious ROI's, eliminate false positives, and classify the remaining as malignant or benign. The focus of attention module uses a difference of Gaussians (DoG) filter to highlight suspicious regions in the mammogram. The index module uses tests to reduce the number of nonmalignant regions from 8.39 to 2.36 per full breast image. Size, shape, contrast, and Laws texture features are used to develop the prediction module's mass models. Derivative-based feature saliency techniques are used to determine the best features for classification. Nine features are chosen to define the malignant/benign models. The feature extraction module obtains these features from all suspicious ROI's. The matching module classifies the regions using a multilayer perceptron neural network architecture to obtain an overall classification accuracy of 100% for the segmented malignant masses with a false-positive rate of 1.8 per full breast image. This system has a sensitivity of 92% for locating malignant ROI's. The database contains 272 images (12 b, 100 μm) with 36 malignant and 53 benign mass images. The results demonstrate that the MBV approach provides a structured order of integrating complex stages into a system for radiologists     

A multiple circular path convolution neural network system for detection of mammographic masses

A multiple circular path convolution neural network (MCPCNN) architecture specifically designed for the analysis of tumor and tumor-like structures has been constructed. We first divided each suspected tumor area into sectors and computed the defined mass features for each sector independently. These sector features were used on the input layer and were coordinated by convolution kernels of different sizes that propagated signals to the second layer in the neural network system. The convolution kernels were trained, as required, by presenting the training cases to the neural network. In this study, randomly selected mammograms were processed by a dual morphological enhancement technique. Radiodense areas were isolated and were delineated using a region growing algorithm. The boundary of each region of interest was then divided into 36 sectors using 36 equi-angular dividers radiated from the center of the region. A total of 144 Breast Imaging-Reporting and Data System-based features (i.e., four features per sector for 36 sectors) were computed as input values for the evaluation of this newly invented neural network system. The overall performance was 0.78-0.80 for the areas (Az) under the receiver operating characteristic curves using the conventional feed-forward neural network in the detection of mammographic masses. The performance was markedly improved with Az values ranging from 0.84 to 0.89 using the MCPCNN. This paper does not intend to claim the best mass detection system. Instead it reports a potentially better neural network structure for analyzing a set of the mass features defined by an investigator.     

Computer-aided diagnosis of solid breast nodules: use of an artificial neural network based on multiple sonographic features

A computer-aided diagnosis (CAD) algorithm identifying breast nodule malignancy using multiple ultrasonography (US) features and artificial neural network (ANN) classifier was developed from a database of 584 histologically confirmed cases containing 300 benign and 284 malignant breast nodules. The features determining whether a breast nodule is benign or malignant were extracted from US images through digital image processing with a relatively simple segmentation algorithm applied to the manually preselected region of interest. An ANN then distinguished malignant nodules in US images based on five morphological features representing the shape, edge characteristics, and darkness of a nodule. The structure of ANN was selected using k-fold cross-validation method with k = 10. The ANN trained with randomly selected half of breast nodule images showed the normalized area under the receiver operating characteristic curve of 0.95. With the trained ANN, 53.3% of biopsies on benign nodules can be avoided with 99.3% sensitivity. Performance of the developed classifier was reexamined with new US mass images in the generalized patient population of total 266 (167 benign and 99 malignant) cases. The developed CAD algorithm has the potential to increase the specificity of US for characterization of breast lesions.     

Automatic identification of the pectoral muscle in mammograms

The pectoral muscle represents a predominant density region in most medio-lateral oblique (MLO) views of mammograms; its inclusion can affect the results of intensity-based image processing methods or bias procedures in the detection of breast cancer. Local analysis of the pectoral muscle may be used to identify the presence of abnormal axillary lymph nodes, which may be the only manifestation of occult breast carcinoma. We propose a new method for the identification of the pectoral muscle in MLO mammograms based upon a multiresolution technique using Gabor wavelets. This new method overcomes the limitation of the straight-line representation considered in our initial investigation using the Hough transform. The method starts by convolving a group of Gabor filters, specially designed for enhancing the pectoral muscle edge, with the region of interest containing the pectoral muscle. After computing the magnitude and phase images using a vector-summation procedure, the magnitude value of each pixel is propagated in the direction of the phase. The resulting image is then used to detect the relevant edges. Finally, a post-processing stage is used to find the true pectoral muscle edge. The method was applied to 84 MLO mammograms from the Mini-MIAS (Mammographic Image Analysis Society, London, U.K.) database. Evaluation of the pectoral muscle edge detected in the mammograms was performed based upon the percentage of false-positive (FP) and false-negative (FN) pixels determined by comparison between the numbers of pixels enclosed in the regions delimited by the edges identified by a radiologist and by the proposed method. The average FP and FN rates were, respectively, 0.58% and 5.77%. Furthermore, the results of the Gabor-filter-based method indicated low Hausdorff distances with respect to the hand-drawn pectoral muscle edges, with the mean and standard deviation being 3.84 +/- 1.73 mm over 84 images.     

An anatomically oriented breast coordinate system for mammogram analysis

We have developed a breast coordinate system that is based on breast anatomy to register female breasts into a common coordinate frame in 2-D mediolateral (ML) or mediolateral oblique (MLO) view mammograms. The breasts are registered according to the location of the pectoral muscle and the nipple and the shape of the breast boundary because these are the most robust features independent of the breast size and shape. On the basis of these landmarks, we have constructed a nonlinear mapping between the parameter frame and the breast region in the mammogram. This mapping makes it possible to identify the corresponding positions and orientations among all of the ML or MLO mammograms, which facilitates an implicit use of the registration, i.e., no explicit image warping is needed. We additionally show how the coordinate transform can be used to extract Gaussian derivative features so that the feature positions and orientations are registered and extracted without nonlinearly deforming the images. We use the proposed breast coordinate transform in a cross-sectional breast cancer risk assessment study of 490 women, in which we attempt to learn breast cancer risk factors from mammograms that were taken prior to when the breast cancer became visible to a radiologist. The coordinate system provides both the relative position and orientation information on the breast region from which the features are derived. In addition, the coordinate system can be used in temporal studies to pinpoint anatomically equivalent locations between the mammograms of each woman and among the mammograms of all of the women in the study. The results of the cross-sectional study show that the classification into cancer and control groups can be improved by using the new coordinate system, compared to other systems evaluated. Comparisons were performed using the area-under-the-receiver-operating-characteristic-curve score. In general, the new coordinate system makes an accurate anatomical registration of breasts possible, which suggests its wide applicability wherever 2-D mammogram registration is required.     

Automated seeded lesion segmentation on digital mammograms

Segmenting lesions is a vital step in many computerized mass-detection schemes for digital (or digitized) mammograms. The authors have developed two novel lesion segmentation techniques-one based on a single feature called the radial gradient index (RGI) and one based on simple probabilistic models to segment mass lesions, or other similar nodular structures, from surrounding background. In both methods a series of image partitions is created using gray-level information as well as prior knowledge of the shape of typical mass lesions. With the former method the partition that maximizes the RGI is selected. In the latter method, probability distributions for gray-levels inside and outside the partitions are estimated, and subsequently used to determine the probability that the image occurred for each given partition. The partition that maximizes this probability is selected as the final lesion partition (contour). The authors tested these methods against a conventional region growing algorithm using a database of biopsy-proven, malignant lesions and found that the new lesion segmentation algorithms more closely match radiologists' outlines of these lesions. At an overlap threshold of 0.30, gray level region growing correctly delineates 62% of the lesions in the authors' database while the RGI and probabilistic segmentation algorithms correctly segment 92% and 96% of the lesions, respectively     

Regional Admittivity Spectra With Tomosynthesis Images for Breast Cancer Detection: Preliminary Patient Study

It has been known for some time that many tumors have a significantly different conductivity and permittivity from surrounding normal tissue. This high “contrast” in tissue electrical properties, occurring between a few kilohertz and several megahertz, may permit differentiating malignant from benign tissues. Here we show the ability of electrical impedance spectroscopy (EIS) to roughly localize and clearly distinguish cancers from normal tissues and benign lesions. Localization of these lesions is confirmed by simultaneous, in register digital breast tomosynthesis (DBT) mammography or 3-D mammograms.      

An approach to automated detection of tumors in mammograms

An automated system for detecting and classifying particular types of tumors in digitized mammograms is described. The analysis of mammograms is performed in two stages. First, the system identifies pixel groupings that may correspond to tumors. Next, detected pixel groupings are subjected to classification. The essence of the first processing stage is multiresolution image processing based on fuzzy pyramid linking. The second stage uses a classification hierarchy to identify benign and malignant tumors. Each level of the hierarchy uses deterministic or Bayes classifiers and a particular measurement. The measurements pertain to shape and intensity characteristics of particular types of tumors. The classification hierarchy is organized in such a way that the simplest measurements are used at the top, with the system stepping through the hierarchy only when it cannot classify the detected pixel groupings with certainty.     

Distributing Charges on a Sphere by means of Evolutionary Strategies

This paper analyzes measures based on texture by means of the Spatial Gray Level Dependence Method, and geometry using nodules skeleton, with the purpose of characterizing solitary lung nodules as malignant or benign in computerized tomography images. Based on a sample of 31 nodules, 25 benign and 6 malignant, these methods are first analyzed individually and then jointly, with classification and analysis techniques (linear stepwise discriminant analysis, leave-one-out and ROC curve). We have concluded that the individual measures and their combinations produce good results in the diagnosis of solitary lung nodules. Nevertheless, there is the need to perform tests with a larger database and more complex cases in order to obtain a more precise behavior pattern.     

Effect of lesion morphology on microwave signature in 2-D ultra-wideband breast imaging

This paper studies the possibility of distinguishing between benign and malignant masses by exploiting the morphology-dependent temporal and spectral characteristics of their microwave backscatter response in ultra-wideband breast cancer detection. The spiculated border profiles of 2-D breast masses are generated by modifying the baseline elliptical rings based upon the irregularity of their peripheries. Furthermore, the single- and multilayer lesion models are used to characterize a distinct mass region followed by a sharp transition to background, and a blurred mass border exhibiting a gradual transition to background, respectively. Subsequently, the complex natural resonances (CNRs) of the backscatter microwave signature can be derived from the late-time target response and reveal diagnostically useful information. The fractional sequence CLEAN algorithm is proposed to estimate the lesions' delay intervals and identify the late-time responses. Finally, it is shown through numerical examples that the locations of dominant CNRs are dependent on the lesion morphologies, where 2-D computational breast phantoms with single and multiple lesions are investigated. The analysis is of potential use for discrimination between benign and malignant lesions, where the former usually possesses a better-defined, more compact shape as opposed to the latter.