Search

In this study, we proposed a multi-space-enabled deep learning modeling method for predicting Oncotype DX recurrence risk categories from digital mammogram images on breast cancer patients. Our study included 189 estrogen receptor-positive (ER+) and node-negative invasive breast cancer patients, who all have Oncotype DX recurrence risk score available. Breast tumors were segmented manually by an expert radiologist. We built a 3- channel convolutional neural network (CNN) model that accepts three-space tumor data: the spatial intensity information and the phase and amplitude components in the frequency domain. We compared this multi-space model to a baseline model that is based on sorely the intensity information. Classification accuracy is based on 5- fold cross-validation and average area-under the receiver operating characteristics curve (AUC). Our results showed that the 3-channel multi-space CNN model achieved a statistically significant improvement than the baseline model.

In order to establish a new personalized breast cancer screening paradigm, it is critically important to accurately predict the short-term risk of a woman having image-detectable cancer after a negative mammographic screening. In this study, we developed and tested a novel short-term risk assessment model based on deep learning method. During the experiment, a number of 270 “prior” negative screening cases was assembled. In the next sequential (“current”) screening mammography, 135 cases were positive and 135 cases remained negative. These cases were randomly divided into a training set with 200 cases and a testing set with 70 cases. A deep learning based computer-aided diagnosis (CAD) scheme was then developed for the risk assessment, which consists of two modules: adaptive feature identification module and risk prediction module. The adaptive feature identification module is composed of three pairs of convolution-max-pooling layers, which contains 20, 10, and 5 feature maps respectively. The risk prediction module is implemented by a multiple layer perception (MLP) classifier, which produces a risk score to predict the likelihood of the woman developing short-term mammography-detectable cancer. The result shows that the new CAD-based risk model yielded a positive predictive value of 69.2% and a negative predictive value of 74.2%, with a total prediction accuracy of 71.4%. This study demonstrated that applying a new deep learning technology may have significant potential to develop a new short-term risk predicting scheme with improved performance in detecting early abnormal symptom from the negative mammograms.

Collection of large repositories of clinical images containing verified cancer locations is costly and time consuming due to difficulties associated with both the accumulation of data and establishment of the ground truth. This problem poses a significant challenge to the development of machine learning algorithms that require large amounts of data to properly train and avoid overfitting. In this paper we expand the methods in our previous publications by making several modifications that significantly increase the speed of our insertion algorithms, thereby allowing them to be used for inserting lesions that are much larger in size. These algorithms have been incorporated into an image composition tool that we have made publicly available. This tool allows users to modify or supplement existing datasets by seamlessly inserting a real breast mass or micro-calcification cluster extracted from a source digital mammogram into a different location on another mammogram. We demonstrate examples of the performance of this tool on clinical cases taken from the University of South Florida Digital Database for Screening Mammography (DDSM). Finally, we report the results of a reader study evaluating the realism of inserted lesions compared to clinical lesions. Analysis of the radiologist scores in the study using receiver operating characteristic (ROC) methodology indicates that inserted lesions cannot be reliably distinguished from clinical lesions.

This paper presents a framework for mammogram enhancement that is based on a selective enhancement technique. Several enhancement algorithms under this framework are developed, which include weighted mean gray value- and fuzzy cross-over point-based thresholding methods, algorithm fusion, iterative enhancement method, and statistical decision theory-based techniques. Using various abnormal mammograms, the presented algorithms prove to be more robust and yield superior performance when compared with six representative enhancement approaches available in the literature.

Microcalcification detection in a mammogram is an effective method to find the early stage of breast tumor. Especially, computer aided diagnosis (CAD) improves the working performance of radiologists and doctors as it offers an efficient microcalcification detection. In this paper, we propose a microcalcification detection system which consists of three modules; coarse detection, clustering, and fine detection module. The coarse detection module finds candidate pixels from an entire mammogram which are suspected as a part of a microcalcification. The module not only extracts two median contrast features and two contrast-to-noise ratio features, but also categorizes the candidate pixels with a linear kernel-based SVM classifier. Then, the clustering module forms the candidate pixels into regions of interest (ROI) using a region growing algorithm. The objective of the fine detection module is to decide whether the corresponding region classifies as a microcalcification or not. Eleven features including distribution, variance, gradient, and various edge components are extracted from the clustered ROIs and are fed into a radial basis function-based SVM classifier to determine the microcalcification. In order to verify the effectiveness of the proposed microcalcification detection system, the experiments are performed with full-field digital mammogram (FFDM). We also compare its detection performance with an ANN-based detection system.

In this paper, a novel wavelet-morphology method for the detection of mass abnormalities in digital mammograms is presented. The new scheme utilizes the feature extraction capability of the wavelet transform followed by a novel recursive-enhancement morphology algorithm to detect the masses. A morphology-based segmentation algorithm is finally applied to the enhanced image to separate the mass from the normal breast tissues. This technique outlines the shape of the region of interest (mass in mammograms). Tests results have confirmed the efficacy of the technique in automated detection of abnormalities in wavelet based compressed mammograms.

This work incorporates development of a model that relates 3-D arrangement of breast tissue and its appearance in a mammogram. The primary contribution is modeling breast structures based on their anatomic properties, obtained from the literature and histologic slice images. Additionally, breast compression deformation is estimated using tissue elastic properties and a deformation model. Synthetic mammograms obtained by the proposed approach can be used in analysis of positioning and compression effects, and for testing computer algorithms for detection of abnormalities. The paper discusses three major model components: (1) modeling breast anatomic structures, (2) modeling breast tissue compression, and (3) modeling X-ray image acquisition.

Eye position of observes representing four levels of training and experience: mammographers; mammography residents; mammographic technologists; and, laypersons were compared to a random search model as they examined a set of nine two-view digital mammogram pairs for breast masses. Analysis of time-to-hit data revealed that mammographers' training and experience combined to produce the most efficient search patterns as measured by the fastest search times to detect breast masses on two views. Scanning patterns of mammography residents and mammographic technologists were less efficient due to wider dispersion of visual attention that was divided between potential breast masses and perturbations in breast parenchyma. Because laypersons lacked training in radiology, bright blobs in the breast image were considered to be intuitively valid target candidates, and these features distracted search by capturing visual attention.© (1996) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.