Detection of Clustered Microcalcifications on Mammograms Using Surrounding Region Dependence Method and Artificial Neural Network

Clustered microcalcifications on X-ray mammograms are an important sign in the detection of breast cancer. A statistical texture analysis method, called the surrounding region dependence method (SRDM), is proposed for the detection of clustered microcalcifications on digitized mammograms. The SRDM is based on the second-order histogram in two surrounding regions. This method defines four textural features to classify region of interests (ROIs) into positive ROIs containing clustered microcalcifications and negative ROIs of normal tissues. The database is composed of 64 positive and 76 negative ROI images, which are selected from digitized mammograms with a pixel size of 100 × 100 m² and 12 bits per pixel. An ROI is selected as an area of 128 × 128 pixels on the digitized mammograms. In order to classify ROIs into the two types, a three-layer backpropagation neural network is employed as a classifier. A segmentation of individual microcalcifications is also proposed to show their morphologies. The classification performance of the proposed method is evaluated by using the round-robin method and a free-response receiver operating-characteristics (FROC) analysis. A receiver operating-characteristics (ROC) analysis is employed to present the results of the round-robin testing for the case of several hidden neurons. The area under the ROC curve, A _z, is 0.997, which is achieved in the case of 4 hidden neurons. The FROC analysis is performed on 20 cropped images. A cropped image is selected as an area of 512 × 512 pixels on the digitized mammograms. In terms of the FROC, a sensitivity of more than 90% is obtained with a low false-positive (FP) detection rate of 0.67 per cropped image.     

A CAD system for the automatic detection of clustered microcalcifications in digitized mammogram films

Clusters of microcalcifications in mammograms are an important early sign of breast cancer. This paper presents a computer-aided diagnosis (CAD) system for the automatic detection of clustered microcalcifications in digitized mammograms. The proposed system consists of two main steps. First, potential microcalcification pixels in the mammograms are segmented out by using mixed features consisting of wavelet features and gray level statistical features, and labeled into potential individual microcalcification objects by their spatial connectivity. Second, individual microcalcifications are detected by using a set of 31 features extracted from the potential individual microcalcification objects. The discriminatory power of these features is analyzed using general regression neural networks via sequential forward and sequential backward selection methods. The classifiers used in these two steps are both multilayer feedforward neural networks. The method is applied to a database of 40 mammograms (Nijmegen database) containing 105 clusters of microcalcifications. A free-response operating characteristics (FROC) curve is used to evaluate the performance. Results show that the proposed system gives quite satisfactory detection performance. In particular, a 90% mean true positive detection rate is achieved at the cost of 0.5 false positive per image when mixed features are used in the first step and 15 features selected by the sequential backward selection method are used in the second step. However, we must be cautious when interpreting the results, since the 20 training samples are also used in the testing step.     

Classification of mass and normal breast tissue: a convolution neural network classifier with spatial domain and texture images

The authors investigated the classification of regions of interest (ROI's) on mammograms as either mass or normal tissue using a convolution neural network (CNN). A CNN is a backpropagation neural network with two-dimensional (2-D) weight kernels that operate on images. A generalized, fast and stable implementation of the CNN was developed. The input images to the CNN were obtained from the ROI's using two techniques. The first technique employed averaging and subsampling. The second technique employed texture feature extraction methods applied to small subregions inside the ROI. Features computed over different subregions were arranged as texture images, which were subsequently used as CNN inputs. The effects of CNN architecture and texture feature parameters on classification accuracy were studied. Receiver operating characteristic (ROC) methodology was used to evaluate the classification accuracy. A data set consisting of 168 ROIs containing biopsy-proven masses and 504 ROI's containing normal breast tissue was extracted from 168 mammograms by radiologists experienced in mammography. This data set was used for training and testing the CNN. With the best combination of CNN architecture and texture feature parameters, the area under the test ROC curve reached 0.87, which corresponded to a true-positive fraction of 90% at a false positive fraction of 31%. The authors' results demonstrate the feasibility of using a CNN for classification of masses and normal tissue on mammograms.     

Application of Artificial Neural Networks to Medical Image Pattern Recognition: Detection of Clustered Microcalcifications on Mammograms and Lung Cancer on Chest Radiographs

Three neural network models were employed to evaluate their performances in the recognition of medical image patterns associated with lung cancer and breast cancer in radiography. The first method was a pattern match neural network. The second was a conventional backpropagation neural network. The third method was a backpropagation trained neocognitron in which the signal propagation is operated with the convolution calculation from one layer to the next. In the convolution neural network (CNN) experiment, several output association methods and trainer imposed driving functions in conjunction with the convolution neural network are proposed for general medical image pattern recognition. An unconventional method of applying rotation and shift invariance is also used to enhance the performance of the neural nets.We have tested these methods for the detection of microcalcifications on mammograms and lung nodules on chest radiographs. Pre-scan methods were previously described in our early publications. The artificial neural networks act as final detection classifiers to determine if a disease pattern is presented on the suspected image area. We found that the convolution neural network, which internally performs feature extraction and classification, achieves the best performance among the three neural network models. These results show that some processing associated with disease feature extraction is a necessary step before a classifier can make an accurate determination.     

Classification of malignant and benign masses based on hybrid ART2LDA approach

A new type of classifier combining an unsupervised and a supervised model was designed and applied to classification of malignant and benign masses on mammograms. The unsupervised model was based on an adaptive resonance theory (ART2) network which clustered the masses into a number of separate classes. The classes were divided into two types: one containing only malignant masses and the other containing a mix of malignant and benign masses. The masses from the malignant classes were classified by ART2. The masses from the mixed classes were input to a supervised linear discriminant classifier (LDA). In this way, some malignant masses were separated and classified by ART2 and the less distinguishable benign and malignant masses were classified by LDA. For the evaluation of classifier performance, 348 regions of interest (ROI's) containing biopsy proven masses (169 benign and 179 malignant) were used. Ten different partitions of training and test groups were randomly generated using an average of 73% of ROI's for training and 27% for testing. Classifier design, including feature selection and weight optimization, was performed with the training group. The test group was kept independent of the training group. The performance of the hybrid classifier was compared to that of an LDA classifier alone and a backpropagation neural network (BPN). Receiver operating characteristics (ROC) analysis was used to evaluate the accuracy of the classifiers. The average area under the ROC curve (A(z)) for the hybrid classifier was 0.81 as compared to 0.78 for the LDA and 0.80 for the BPN. The partial areas above a true positive fraction of 0.9 were 0.34, 0.27 and 0.31 for the hybrid, the LDA and the BPN classifier, respectively. These results indicate that the hybrid classifier is a promising approach for improving the accuracy of classification in CAD applications.     

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).     

Analysis of mammographic microcalcifications using gray-level image structure features

At present, mammography associated with clinical breast examination and breast self-examination is the only effective and viable method for mass breast screening. The presence of microcalcifications is one of the primary signs of breast cancer. It is, difficult however, to distinguish between benign and malignant microcalcifications associated with breast cancer. Here, the authors define a set of image structure features for classification of malignancy. Two categories of correlated gray-level image structure features are defined for classification of "difficult-to-diagnose" cases. The first category of features includes second-order histogram statistics-based features representing the global texture and the wavelet decomposition-based features representing the local texture of the microcalcification area of interest. The second category of features represents the first-order gray-level histogram-based statistics of the segmented microcalcification regions and the size, number, and distance features of the segmented microcalcification cluster. Various features in each category were correlated with the biopsy examination results of 191 "difficult-to-diagnose" cases for selection of the best set of features representing the complete gray-level image structure information. The selection of the best features was performed using the multivariate cluster analysis as well as a genetic algorithm (GA)-based search method. The selected features were used for classification using backpropagation neural network and parameteric statistical classifiers. Receiver operating characteristic (ROC) analysis was performed to compare the neural network-based classification with linear and k-nearest neighbor (KNN) classifiers. The neural network classifier yielded better results using the combined set of features selected through the GA-based search method for classification of "difficult-to-diagnose" microcalcifications.     

Detection of microcalcifications in digital mammograms usingwavelets

This paper presents an approach for detecting micro-calcifications in digital mammograms employing wavelet-based subband image decomposition. The microcalcifications appear in small clusters of few pixels with relatively high intensity compared with their neighboring pixels. These image features can be preserved by a detection system that employs a suitable image transform which can localize the signal characteristics in the original and the transform domain. Given that the microcalcifications correspond to high-frequency components of the image spectrum, detection of microcalcifications is achieved by decomposing the mammograms into different frequency subbands, suppressing the low-frequency subband, and, finally, reconstructing the mammogram from the subbands containing only high frequencies. Preliminary experiments indicate that further studies are needed to investigate the potential of wavelet-based subband image decomposition as a tool for detecting microcalcifications in digital mammograms     

Integrated wavelets for enhancement of microcalcifications in digital mammography

This paper presents a new algorithm for enhancement of microcalcifications in mammograms. The main novelty is the application of techniques we have developed for construction of filterbanks derived from the continuous wavelet transform. These discrete wavelet decompositions, called integrated wavelets, are optimally designed for enhancement of multiscale structures in images. Furthermore, we use a model based approach to refine existing methods for general enhancement of mammograms resulting in a more specific enhancement of microcalcifications. We present results of our method and compare them with known algorithms. Finally, we want to indicate how these techniques can also be applied to the detection of microcalcifications. Our algorithm was positively evaluated in a clinical study. It has been implemented in a mammography workstation designed for soft-copy reading of digital mammograms developed by IMAGETOOL, Germany.     

Optimal filter-based detection of microcalcifications

This paper deals with the problem of texture feature extraction in digital mammograms. We use the extracted features to discriminate between texture representing clusters of microcalcifications and texture representing normal tissue. Having a two-class problem, we suggest a texture feature extraction method based on a single filter optimized with respect to the Fisher criterion. The advantage of this criterion is that it uses both the feature mean and the feature variance to achieve good feature separation. Image compression is desirable to facilitate electronic transmission and storage of digitized mammograms. In this paper, we also explore the effects of data compression on the performance of our proposed detection scheme. The mammograms in our test set were compressed at different ratios using the Joint Photographic Experts Group compression method. Results from an experimental study indicate that our scheme is very well suited for detecting clustered microcalcifications in both uncompressed and compressed mammograms. For the uncompressed mammograms, at a rate of 1.5 false positive clusters/image our method reaches a true positive rate of about 95%, which is comparable to the best results achieved so far. The detection performance for images compressed by a factor of about four is very similar to the performance for uncompressed images.     

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     

Scale-space signatures for the detection of clustered microcalculations in digital mammograms

A method is described for the automated detection of microcalcifications in digitized mammograms. The method is based on the Laplacian scale-space representation of the mammogram only. First, possible locations of microcalcifications are identified as local maxima in the filtered image on a range of scales. For each finding, the size and local contrast is estimated, based on the Laplacian response denoted as the scale-space signature. A finding is marked as a microcalcification if the estimated contrast is larger than a predefined threshold which depends on the size of the finding. It is shown that the signature has a characteristic peak, revealing the corresponding image features. This peak can be robustly determined. The basic method is significantly improved by consideration of the statistical variation of the estimated contrast, which is the result of the complex noise characteristic of the mammograms. The method is evaluated with the Nijmegen database and compared to other methods using these mammograms. Results are presented as the free-response receiver operating characteristic (FROC) performance. At a rate of one false positive cluster per image the method reaches a sensitivity of 0.84, which is comparable to the best results achieved so far.     

Optimization of wavelet decomposition for image compression and feature preservation

A neural-network-based framework has been developed to search for an optimal wavelet kernel that can be used for a specific image processing task. In this paper, a linear convolution neural network was employed to seek a wavelet that minimizes errors and maximizes compression efficiency for an image or a defined image pattern such as microcalcifications in mammograms and bone in computed tomography (CT) head images. We have used this method to evaluate the performance of tap-4 wavelets on mammograms, CTs, magnetic resonance images, and Lena images. We found that the Daubechies wavelet or those wavelets with similar filtering characteristics can produce the highest compression efficiency with the smallest mean-square-error for many image patterns including general image textures as well as microcalcifications in digital mammograms. However, the Haar wavelet produces the best results on sharp edges and low-noise smooth areas. We also found that a special wavelet whose low-pass filter coefficients are 0.32252136, 0.85258927, 1.38458542, and -0.14548269) produces the best preservation outcomes in all tested microcalcification features including the peak signal-to-noise ratio, the contrast and the figure of merit in the wavelet lossy compression scheme. Having analyzed the spectrum of the wavelet filters, we can find the compression outcomes and feature preservation characteristics as a function of wavelets. This newly developed optimization approach can be generalized to other image analysis applications where a wavelet decomposition is employed.     

Wavelet transforms for detecting microcalcifications in mammograms

Clusters of fine, granular microcalcifications in mammograms may be an early sign of disease. Individual grains are difficult to detect and segment due to size and shape variability and because the background mammogram texture is typically inhomogeneous. The authors develop a 2-stage method based on wavelet transforms for detecting and segmenting calcifications. The first stage is based on an undecimated wavelet transform, which is simply the conventional filter bank implementation without downsampling, so that the low-low (LL), low-high (LH), high-low (HL), and high-high (HH) sub-bands remain at full size. Detection takes place in HH and the combination LH+HL. Four octaves are computed with 2 inter-octave voices for finer scale resolution. By appropriate selection of the wavelet basis the detection of microcalcifications in the relevant size range can be nearly optimized. In fact, the filters which transform the input image into HH and LH+HL are closely related to prewhitening matched filters for detecting Gaussian objects (idealized microcalcifications) in 2 common forms of Markov (background) noise. The second stage is designed to overcome the limitations of the simplistic Gaussian assumption and provides an accurate segmentation of calcification boundaries. Detected pixel sites in HH and LH+HL are dilated then weighted before computing the inverse wavelet transform. Individual microcalcifications are greatly enhanced in the output image, to the point where straightforward thresholding can be applied to segment them. FROG curves are computed from tests using a freely distributed database of digitized mammograms.     

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     

Adaptive mammographic image enhancement using first derivative andlocal statistics

This paper proposes an adaptive image enhancement method for mammographic images, which is based on the first derivative and the local statistics. The adaptive enhancement method consists of three processing steps. The first step is to remove the film artifacts which may be misread as microcalcifications. The second step is to compute the gradient images by using the first derivative operators. The third step is to enhance the important features of the mammographic image by adding the adaptively weighted gradient images. Local statistics of the image are utilized for adaptive realization of the enhancement, so that image details can be enhanced and image noises can be suppressed. The objective performances of the proposed method were compared with those by the conventional image enhancement methods for a simulated image and the seven mammographic images containing real microcalcifications. The performance of the proposed method was also evaluated by means of the receiver operating characteristics (ROC) analysis for 78 real mammographic images with and without microcalcifications     

A similarity learning approach to content-based image retrieval: application to digital mammography

In this paper, we describe an approach to content-based retrieval of medical images from a database, and provide a preliminary demonstration of our approach as applied to retrieval of digital mammograms. Content-based image retrieval (CBIR) refers to the retrieval of images from a database using information derived from the images themselves, rather than solely from accompanying text indices. In the medical-imaging context, the ultimate aim of CBIR is to provide radiologists with a diagnostic aid in the form of a display of relevant past cases, along with proven pathology and other suitable information. CBIR may also be useful as a training tool for medical students and residents. The goal of information retrieval is to recall from a database information that is relevant to the user's query. The most challenging aspect of CBIR is the definition of relevance (similarity), which is used to guide the retrieval machine. In this paper, we pursue a new approach, in which similarity is learned from training examples provided by human observers. Specifically, we explore the use of neural networks and support vector machines to predict the user's notion of similarity. Within this framework we propose using a hierarchal learning approach, which consists of a cascade of a binary classifier and a regression module to optimize retrieval effectiveness and efficiency. We also explore how to incorporate online human interaction to achieve relevance feedback in this learning framework. Our experiments are based on a database consisting of 76 mammograms, all of which contain clustered microcalcifications (MCs). Our goal is to retrieve mammogram images containing similar MC clusters to that in a query. The performance of the retrieval system is evaluated using precision-recall curves computed using a cross-validation procedure. Our experimental results demonstrate that: 1) the learning framework can accurately predict the perceptual similarity reported by human observers, thereby serving as a basis for CBIR; 2) the learning-based framework can significantly outperform a simple distance-based similarity metric; 3) the use of the hierarchical two-stage network can improve retrieval performance; and 4) relevance feedback can be effectively incorporated into this learning framework to achieve improvement in retrieval precision based on online interaction with users; and 5) the retrieved images by the network can have predicting value for the disease condition of the query.     

A novel approach to microcalcification detection using fuzzy logictechnique

Breast cancer continues to be a significant public health problem in the United States. Approximately, 182,000 new cases of breast cancer are diagnosed and 46,000 women die of breast cancer each year. Even more disturbing is the fact that one out of eight women in the United States will develop breast cancer at some point during her lifetime. Since the cause of breast cancer remains unknown, primary prevention becomes impossible. Computer-aided mammography is an important and challenging task in automated diagnosis. It has great potential over traditional interpretation of film-screen mammography in terms of efficiency and accuracy. Microcalcifications are the earliest sign of breast carcinomas and their detection is one of the key issues for breast cancer control. In this study, a novel approach to microcalcification detection based on fuzzy logic technique is presented. Microcalcifications are first enhanced based on their brightness and nonuniformity. Then, the irrelevant breast structures are excluded by a curve detector. Finally, microcalcifications are located using an iterative threshold selection method. The shapes of microcalcifications are reconstructed and the isolated pixels are removed by employing the mathematical morphology technique. The essential idea of the proposed approach is to apply a fuzzified image of a mammogram to locate the suspicious regions and to interact the fuzzified image with the original image to preserve fidelity. The major advantage of the proposed method is its ability to detect microcalcifications even in very dense breast mammograms. A series of clinical mammograms are employed to test the proposed algorithm and the performance is evaluated by the free-response receiver operating characteristic curve. The experiments aptly show that the microcalcifications can be accurately detected even in very dense mammograms using the proposed approach     

Optimal linear transformation for MRI feature extraction

This paper presents development and application of a feature extraction method for magnetic resonance imaging (MRI), without explicit calculation of tissue parameters. A three-dimensional (3-D) feature space representation of the data is generated in which normal tissues are clustered around prespecified target positions and abnormalities are clustered elsewhere. This is accomplished by a linear minimum mean square error transformation of categorical data to target positions. From the 3-D histogram (cluster plot) of the transformed data, clusters are identified and regions of interest (ROI's) for normal and abnormal tissues are defined. These ROI's are used to estimate signature (prototype) vectors for each tissue type which in turn are used to segment the MRI scene. The proposed feature space is compared to those generated by tissue-parameter-weighted images, principal component images, and angle images, demonstrating its superiority for feature extraction and scene segmentation. Its relationship with discriminant analysis is discussed. The method and its performance are illustrated using a computer simulation and MRI images of an egg phantom and a human brain.     

Relevance vector machine for automatic detection of clustered microcalcifications

Clustered microcalcifications (MC) in mammograms can be an important early sign of breast cancer in women. Their accurate detection is important in computer-aided detection (CADe). In this paper, we propose the use of a recently developed machine-learning technique--relevance vector machine (RVM)--for detection of MCs in digital mammograms. RVM is based on Bayesian estimation theory, of which a distinctive feature is that it can yield a sparse decision function that is defined by only a very small number of so-called relevance vectors. By exploiting this sparse property of the RVM, we develop computerized detection algorithms that are not only accurate but also computationally efficient for MC detection in mammograms. We formulate MC detection as a supervised-learning problem, and apply RVM as a classifier to determine at each location in the mammogram if an MC object is present or not. To increase the computation speed further, we develop a two-stage classification network, in which a computationally much simpler linear RVM classifier is applied first to quickly eliminate the overwhelming majority, non-MC pixels in a mammogram from any further consideration. The proposed method is evaluated using a database of 141 clinical mammograms (all containing MCs), and compared with a well-tested support vector machine (SVM) classifier. The detection performance is evaluated using free-response receiver operating characteristic (FROC) curves. It is demonstrated in our experiments that the RVM classifier could greatly reduce the computational complexity of the SVM while maintaining its best detection accuracy. In particular, the two-stage RVM approach could reduce the detection time from 250 s for SVM to 7.26 s for a mammogram (nearly 35-fold reduction). Thus, the proposed RVM classifier is more advantageous for real-time processing of MC clusters in mammograms.