Computerized detection of breast masses in digitized mammograms

We propose a system to detect malignant masses on mammograms. We investigated the behavior of an iris filter at different scales. After iris filter was applied, suspicious regions were segmented by means of an adaptive threshold. Suspected regions were characterized with features based on the iris filter output and, gray level, texture, contour-related, and morphological features extracted from the image. A backpropagation neural network classifier was trained to reduce the number of false positives. The system was developed and evaluated with two completely independent data sets. Results for a test set of 66 malignant and 49 normal cases, evaluated with free-response receiver operating characteristic analysis, yielded a sensitivity of 88% and 94% at 1.02 false positives per image for lesion-based and case-based evaluation, respectively. Results suggest that the proposed method could help radiologists as a second reader in mammographic screening.     

A volumetric method for estimation of breast density on digitized screen-film mammograms

A method is described for the quantitative volumetric analysis of the mammographic density (VBD) from digitized screen-film mammograms. The method is based on initial calibration of the imaging system with a tissue-equivalent plastic device and the subsequent correction for variations in exposure factors and film processing characteristics through images of an aluminum step wedge placed adjacent to the breast during imaging. From information about the compressed breast thickness and technique factors used for taking the mammogram as well as the information from the calibration device, VBD is calculated. First, optical sensitometry is used to convert images to Log relative exposure. Second, the images are corrected for x-ray field inhomogeneity using a spherical section PMMA phantom image. The effectiveness of using the aluminum step wedge in tracking down the variations in exposure factors and film processing was tested by taking test images of the calibration device, aluminum step wedge and known density phantoms at various exposure conditions and also at different times over one year. Results obtained on known density phantoms show that VBD can be estimated to within 5% accuracy from the actual value. A first order thickness correction is employed to correct for inaccuracy in the compression thickness indicator of the mammography units. Clinical studies are ongoing to evaluate whether VBD can be a better indicator for breast cancer risk.     

Computer-aided detection system for clustered microcalcifications: comparison of performance on full-field digital mammograms and digitized screen-film mammograms

We have developed a computer-aided detection (CAD) system to detect clustered microcalcifications automatically on full-field digital mammograms (FFDMs) and a CAD system for screen-film mammograms (SFMs). The two systems used the same computer vision algorithms but their false positive (FP) classifiers were trained separately with sample images of each modality. In this study, we compared the performance of the CAD systems for detection of clustered microcalcifications on pairs of FFDM and SFM obtained from the same patient. For case-based performance evaluation, the FFDM CAD system achieved detection sensitivities of 70%, 80% and 90% at an average FP cluster rate of 0.07, 0.16 and 0.63 per image, compared with an average FP cluster rate of 0.15, 0.38 and 2.02 per image for the SFM CAD system. The difference was statistically significant with the alternative free-response receiver operating characteristic (AFROC) analysis. When evaluated on data sets negative for microcalcification clusters, the average FP cluster rates of the FFDM CAD system were 0.04, 0.11 and 0.33 per image at detection sensitivity level of 70%, 80% and 90% compared with an average FP cluster rate of 0.08, 0.14 and 0.50 per image for the SFM CAD system. When evaluated for malignant cases only, the difference of the performance of the two CAD systems was not statistically significant with AFROC analysis.     

Quantitative assessment of breast density from digitized mammograms into Tabar's patterns

We describe a semi-automated technique for the quantitative assessment of breast density from digitized mammograms in comparison with patterns suggested by Tabar. It was developed using the MATLAB-based graphical user interface applications. It is based on an interactive thresholding method, after a short automated method that shows the fibroglandular tissue area, breast area and breast density each time new thresholds are placed on the image. The breast density is taken as a percentage of the fibroglandular tissue to the breast tissue areas. It was tested in four different ways, namely by examining: (i) correlation of the quantitative assessment results with subjective classification, (ii) classification performance using the quantitative assessment technique, (iii) interobserver agreement and (iv) intraobserver agreement. The results of the quantitative assessment correlated well (r2 = 0.92) with the subjective Tabar patterns classified by the radiologist (correctly classified 83% of digitized mammograms). The average kappa coefficient for the agreement between the readers was 0.63. This indicated moderate agreement between the three observers in classifying breast density using the quantitative assessment technique. The kappa coefficient of 0.75 for intraobserver agreement reflected good agreement between two sets of readings. The technique may be useful as a supplement to the radiologist's assessment in classifying mammograms into Tabar's pattern associated with breast cancer risk.     

Knowledge-based computer-aided detection of masses on digitized mammograms: a preliminary assessment

The purpose of this work was to develop and evaluate a computer-aided detection (CAD) scheme for the improvement of mass identification on digitized mammograms using a knowledge-based approach. Three hundred pathologically verified masses and 300 negative, but suspicious, regions, as initially identified by a rule-based CAD scheme, were randomly selected from a large clinical database for development purposes. In addition, 500 different positive and 500 negative regions were used to test the scheme. This suspicious region pruning scheme includes a learning process to establish a knowledge base that is then used to determine whether a previously identified suspicious region is likely to depict a true mass. This is accomplished by quantitatively characterizing the set of known masses, measuring "similarity" between a suspicious region and a "known" mass, then deriving a composite "likelihood" measure based on all "known" masses to determine the state of the suspicious region. To assess the performance of this method, receiver-operating characteristic (ROC) analyses were employed. Using a leave-one-out validation method with the development set of 600 regions, the knowledge-based CAD scheme achieved an area under the ROC curve of 0.83. Fifty-one percent of the previously identified false-positive regions were eliminated, while maintaining 90% sensitivity. During testing of the 1,000 independent regions, an area under the ROC curve as high as 0.80 was achieved. Knowledge-based approaches can yield a significant reduction in false-positive detections while maintaining reasonable sensitivity. This approach has the potential of improving the performance of other rule-based CAD schemes.     

Field nonuniformity correction for quantitative analysis of digitized mammograms

Several factors, including the heel effect, variation in distance from the x-ray source to points in the image and path obliquity contribute to the signal nonuniformity of mammograms. To best use digitized mammograms for quantitative image analysis, these field non-uniformities must be corrected. An empirically based correction method, which uses a bowl-shaped calibration phantom, has been developed. Due to the annular spherical shape of the phantom, its attenuation is constant over the entire image. Remaining nonuniformities are due only to the heel and inverse square effects as well as the variable path through the beam filter, compression plate and image receptor. In logarithmic space, a normalized image of the phantom can be added to mammograms to correct for these effects. Then, an analytical correction for path obliquity in the breast can be applied to the images. It was found that the correction causes the errors associated with field nonuniformity to be reduced from 14% to 2% for a 4 cm block of material corresponding to a combination of 50% fibroglandular and 50% fatty breast tissue. A repeatability study has been conducted to show that in regions as far as 20 cm away from the chest wall, variations due to imaging conditions and phantom alignment contribute to <2% of overall corrected signal.     

Combined adaptive enhancement and region-growing segmentation of breast masses on digitized mammograms.

As an ongoing effort to develop a computer aid for detection of masses on mammograms, we recently designed an object-based region-growing technique to improve mass segmentation. This segmentation method utilizes the density-weighted contrast enhancement (DWCE) filter as a pre-processing step. The DWCE filter adaptively enhances the contrast between the breast structures and the background. Object-based region growing was then applied to each of the identified structures. The region-growing technique uses gray-scale and gradient information to adjust the initial object borders and to reduce merging between adjacent or overlapping structures. Each object is then classified as a breast mass or normal tissue based on extracted morphological and texture features. In this study we evaluated the sensitivity of this combined segmentation scheme and its ability to reduce false positive (FP) detections on a data set of 253 digitized mammograms, each of which contained a biopsy-proven breast mass. It was found that the segmentation scheme detected 98% of the 253 biopsy-proven breast masses in our data set. After final FP reduction, the detection resulted in 4.2 FP per image at a 90% true positive (TP) fraction and 2.0 FPs per image at an 80% TP fraction. The combined DWCE and object-based region growing technique increased the initial detection sensitivity, reduced merging between neighboring structures, and reduced the number of FP detections in our automated breast mass detection scheme.     

A regional registration technique for automated interval change analysis of breast lesions on mammograms

Analysis of interval change is a useful technique for detection of abnormalities in mammographic interpretation. Interval change analysis is routinely used by radiologists and its importance is well-established in clinical practice. As a first step to develop a computerized method for interval change analysis on mammograms, we are developing an automated regional registration technique to identify corresponding lesions on temporal pairs of mammograms. In this technique, the breast is first segmented from the background on the current and previous mammograms. The breast edges are then aligned using a global alignment procedure based on the mutual information between the breast regions in the two images. Using the nipple location and the breast centroid estimated independently on both mammograms, a polar coordinate system is defined for each image. The polar coordinate of the centroid of a lesion detected on the most recent mammogram is used to obtain an initial estimate of its location on the previous mammogram and to define a fan-shaped search region. A search for a matching structure to the lesion is then performed in the fan-shaped region on the previous mammogram to obtain a final estimate of its location. In this study, a quantitative evaluation of registration accuracy has been performed with a data set of 74 temporal pairs of mammograms and ground-truth correspondence information provided by an experienced radiologist. The most recent mammogram of each temporal pair exhibited a biopsy-proven mass. We have investigated the usefulness of correlation and mutual information as search criteria for determining corresponding regions on mammograms for the biopsy-proven masses. In 85% of the cases (63/74 temporal pairs) the region on the previous mammogram that corresponded to the mass on the current mammogram was correctly identified. The region centroid identified by the registration technique had an average distance of 2.8+/-1.9 mm from the centroid of the radiologist-identified region. These results indicate that our new registration technique may be useful for establishing correspondence between structures on current and previous mammograms. Once such a correspondence is established an interval change analysis could be performed to aid in both detection as well as classification of abnormal breast densities.     

A method to test the reproducibility and to improve performance of computer-aided detection schemes for digitized mammograms

The purpose of this study is to develop a new method for assessment of the reproducibility of computer-aided detection (CAD) schemes for digitized mammograms and to evaluate the possibility of using the implemented approach for improving CAD performance. Two thousand digitized mammograms (representing 500 cases) with 300 depicted verified masses were selected in the study. Series of images were generated for each digitized image by resampling after a series of slight image rotations. A CAD scheme developed in our laboratory was applied to all images to detect suspicious mass regions. We evaluated the reproducibility of the scheme using the detection sensitivity and false-positive rates for the original and resampled images. We also explored the possibility of improving CAD performance using three methods of combining results from the original and resampled images, including simple grouping, averaging output scores, and averaging output scores after grouping. The CAD scheme generated a detection score (from 0 to 1) for each identified suspicious region. A region with a detection score >0.5 was considered as positive. The CAD scheme detected 238 masses (79.3% case-based sensitivity) and identified 1093 false-positive regions (average 0.55 per image) in the original image dataset. In eleven repeated tests using original and ten sets of rotated and resampled images, the scheme detected a maximum of 271 masses and identified as many as 2359 false-positive regions. Two hundred and eighteen masses (80.4%) and 618 false-positive regions (26.2%) were detected in all 11 sets of images. Combining detection results improved reproducibility and the overall CAD performance. In the range of an average false-positive detection rate between 0.5 and 1 per image, the sensitivity of the scheme could be increased approximately 5% after averaging the scores of the regions detected in at least four images. At low false-positive rate (e.g., < or =average 0.3 per image), the grouping method alone could increase CAD sensitivity by 7%. The study demonstrated that reproducibility of a CAD scheme can be tested using a set of slightly rotated and resampled images. Because the reproducibility of true-positive detections is generally higher than that of false-positive detections, combining detection results generated from subsets of rotated and resampled images could improve both reproducibility and overall performance of CAD schemes.     

Identification of the breast boundary in mammograms using active contour models

A method for the identification of the breast boundary in mammograms is presented. The method can be used in the preprocessing stage of a system for computeraided diagnosis (CAD) of breast cancer and also in the reduction of image file size in picture archiving and communication system applications. The method started with modification of the contrast of the original image. A binarisation procedure was then applied to the image, and the chain-code algorithm was used to find an approximate breast contour. Finally, the identification of the true breast boundary was performed by using the approximate contour as the input to an active contour model algorithm specially tailored for this purpose. After demarcation of the breast boundary, all artifacts outside the breast region were eliminated. The method was applied to 84 medio-lateral oblique mammograms from the Mini-MIAS database. Evaluation of the detected breast boundary was performed based upon the percentage of false-positive and false-negative pixels determined by a quantitative comparison between the contours identified by a radiologist and those identified by the proposed method. The average false positive and false negative rates were 0.41% and 0.58%, respectively. The two radiologists who evaluated the results considered the segmentation results to be acceptable for CAD purposes.     

基于TensorMixup的脑胶质瘤全自动分割

脑胶质瘤及其子区域的全自动分割对临床脑胶质瘤患者的诊断、治疗与病情监控具有重要意义。本文改进传统Mixup方法,提出TensorMixup模型,并将其应用于三维U-Net脑胶质瘤分割任务。算法核心思想包括,首先从两位患者相同模态的核磁共振脑影像中分别获取肿瘤区域所在边界框的图像序列,并从获取的图像序列中选取尺寸为128×128×128体素的图像块,然后使用一个所有元素均独立采样于贝塔分布的张量,混合图像块的信息,接着将上述张量映射为矩阵,用于混合图像块的独热编码标签序列,从而合成新图像及其标注数据,最后使用合成数据训练模型,以提高模型的分割精度。在BraTs2019数据集的测试结果显示,本文算法在完整肿瘤、肿瘤核心与增强肿瘤区域的平均Dice值依次可达91.32%、85.67%与82.20%,证明使用TensorMixup进行脑胶质瘤分割,具有可行性与有效性。     

Computer-aided detection of breast masses on full field digital mammograms

We are developing a computer-aided detection (CAD) system for breast masses on full field digital mammographic (FFDM) images. To develop a CAD system that is independent of the FFDM manufacturer's proprietary preprocessing methods, we used the raw FFDM image as input and developed a multiresolution preprocessing scheme for image enhancement. A two-stage prescreening method that combines gradient field analysis with gray level information was developed to identify mass candidates on the processed images. The suspicious structure in each identified region was extracted by clustering-based region growing. Morphological and spatial gray-level dependence texture features were extracted for each suspicious object. Stepwise linear discriminant analysis (LDA) with simplex optimization was used to select the most useful features. Finally, rule-based and LDA classifiers were designed to differentiate masses from normal tissues. Two data sets were collected: a mass data set containing 110 cases of two-view mammograms with a total of 220 images, and a no-mass data set containing 90 cases of two-view mammograms with a total of 180 images. All cases were acquired with a GE Senographe 2000D FFDM system. The true locations of the masses were identified by an experienced radiologist. Free-response receiver operating characteristic analysis was used to evaluate the performance of the CAD system. It was found that our CAD system achieved a case-based sensitivity of 70%, 80%, and 90% at 0.72, 1.08, and 1.82 false positive (FP) marks/image on the mass data set. The FP rates on the no-mass data set were 0.85, 1.31, and 2.14 FP marks/image, respectively, at the corresponding sensitivities. This study demonstrated the usefulness of our CAD techniques for automated detection of masses on FFDM images.     

Dual system approach to computer-aided detection of breast masses on mammograms

In this study, our purpose was to improve the performance of our mass detection system by using a new dual system approach which combines a computer-added detection (CAD) system optimized with "average" masses with another CAD system optimized with "subtle" masses. The two single CAD systems have similar image processing steps, which include prescreening, object segmentation, morphological and texture feature extraction, and false positive (FP) reduction by rule-based and linear discriminant analysis (LDA) classifiers. A feed-forward backpropagation artificial neural network was trained to merge the scores from the LDA classifiers in the two single CAD systems and differentiate true masses from normal tissue. For an unknown test mammogram, the two single CAD systems are applied to the image in parallel to detect suspicious objects. A total of three data sets were used for training and testing the systems. The first data set of 230 current mammograms, referred to as the average mass set, was collected from 115 patients. We also collected 264 mammograms, referred to as the subtle mass set, which were one to two years prior to the current exam from these patients. Both the average and the subtle mass sets were partitioned into two independent data sets in a cross validation training and testing scheme. A third data set containing 65 cases with 260 normal mammograms was used to estimate the FP marker rates during testing. When the single CAD system trained on the average mass set was applied to the test set with average masses, the FP marker rates were 2.2, 1.8, and 1.5 per image at the case-based sensitivities of 90%, 85%, and 80%, respectively. With the dual CAD system, the FP marker rates were reduced to 1.2, 0.9, and 0.7 per image, respectively, at the same case-based sensitivities. Statistically significant (p < 0.05) improvements on the free response receiver operating characteristic curves were observed when the dual system and the single system were compared using the test sets with either average masses or subtle masses.     

A breast density index for digital mammograms based on radiologists’ randing

The purpose of this study was to develop and evaluate a computerized method of calculating a breast density index (BDI) from digitized mammograms that was designed specifically to model radiologists’ perception of breast density. A set of 153 pairs of digitized mammograms (cranio-caudal, CC, and mediolateral oblique, MLO, views) were acquired and preprocessed to reduce detector biases. The sets of mammograms were ordered on an ordinal scale (a scale based only on relative rank-ordering) by two radiologists, and a cardinal (an absolute numerical score) BDI value was calculated from the oridinal ranks. The images were also assigned cardinal BDI values by the radiologists in a subsequent session. Six. mathematical features (including fractal dimension and others) were calculated from the digital mammograms, and were used in conjunction with single value decomposition and multiple linear regression to calculate a computerized BDI. The linear correlation coefficient between different ordinal ranking sessions were as follows: intraradiologist intraprojection (CC/CC):r=0.978; intraradiologist interprojection (CC/MLO):r=0.960; and interradiologist intraprojection (CC/CC):r=0.968. A separate breast density index was derived from three separate ordinal rankings by one radiologist (two with CC views, one with the MLO view). The computer derivedBDI had a correlation coefficient (r) of 0.907 with the radiologists’ ordinalBDI. A comparison between radiologists using a cardinal scoring system (which is closest to how radiologists actually evaluate breast density) showedr=0.914. A breast density index calculated by a computer but modeled after radiologist perception of breast density may be valuable in objectively measuring breast density. Such a metric may prove valuable in numerous areas, including breast cancer risk assessment and in evaluating screening techniques specifically designed to improve imaging of the dense breast.     

Novel network architecture and learning algorithm for the classification of mass abnormalities in digitized mammograms

OBJECTIVE: The main objective of this paper is to present a novel learning algorithm for the classification of mass abnormalities in digitized mammograms. METHODS AND MATERIAL: The proposed approach consists of new network architecture and a new learning algorithm. The original idea is based on the introduction of an additional neuron in the hidden layer for each output class. The additional neurons for benign and malignant classes help in improving memorization ability without destroying the generalization ability of the network. The training is conducted by combining minimal distance-based similarity/random weights and direct calculation of output weights. RESULTS: The proposed approach can memorize training patterns with 100% retrieval accuracy as well as achieve high generalization accuracy for patterns which it has never seen before. The grey-level and breast imaging reporting and data system-based features from digitized mammograms are extracted and used to train the network with the proposed architecture and learning algorithm. The best results achieved by using the proposed approach are 100% on training set and 94% on test set. CONCLUSION: The proposed approach produced very promising results. It has outperformed existing classification approaches in terms of classification accuracy, generalization and memorization abilities, number of iterations, and guaranteed training on a benchmark database.     

Bisphosphonate-functionalized gold nanoparticles for contrast-enhanced X-ray detection of breast microcalcifications

Microcalcifications are one of the most common abnormalities detected by mammography for the diagnosis of breast cancer. However, the detection of microcalcifications and correct diagnosis of breast cancer are limited by the sensitivity and specificity of mammography. Therefore, the objective of this study was to investigate the potential of bisphosphonate-functionalized gold nanoparticles (BP-Au NPs) for contrast-enhanced radiographic detection of breast microcalcifications using two models of breast microcalcifications, which allowed for precise control over levels of hydroxyapatite (HA) mineral within a low attenuating matrix. First, an in vitro imaging phantom was prepared with varying concentrations of HA uniformly dispersed in an agarose hydrogel. The X-ray attenuation of HA-agarose compositions labeled by BP-Au NPs was increased by up to 26 HU compared to unlabeled compositions for HA concentrations ranging from 1 to 10 mg/mL. Second, an ex vivo tissue model was developed to more closely mimic the heterogeneity of breast tissue by injecting varying concentrations of HA in a Matrigel carrier into murine mammary glands. The X-ray attenuation of HA-Matrigel compositions labeled by BP-Au NPs was increased by up to 289 HU compared to unlabeled compositions for HA concentrations ranging from 0.5 to 25 mg/mL, which included an HA concentration (0.5 mg/mL) that was otherwise undetectable by micro-computed tomography. Cumulatively, both models demonstrated the ability of BP-Au NPs to enhance contrast for radiographic detection of microcalcifications, including at a clinically-relevant imaging resolution. Therefore, BP-Au NPs may have potential to improve clinical detection of breast microcalcifications by mammography.     

The effect of intensity windowing on the detection of simulated masses embedded in dense portions of digitized mammograms in a laboratory setting

The purpose of this study was to determine whether intensity windowing (IW) improves detection of simulated masses in dense mammograms. Simulated masses were embedded in dense mammograms digitized at 50 microns/pixel, 12 bits deep. Images were printed with no windowing applied and with nine window width and level combinations applied. A simulated mass was embedded in a realistic background of dense breast tissue, with the position of the mass (against the background) varied. The key variables involved in each trial included the position of the mass, the contrast levels and the IW setting applied to the image. Combining the 10 image processing conditions, 4 contrast levels, and 4 quadrant positions gave 160 combinations. The trials were constructed by pairing 160 combinations of key variables with 160 backgrounds. The entire experiment consisted of 800 trials. Twenty observers were asked to detect the quadrant of the image into which the mass was located. There was a statistically significant improvement in detection performance for masses when the window width was set at 1024 with a level of 3328. IW should be tested in the clinic to determine whether mass detection performance in real mammograms is improved.     

A methodology for evaluation of boundary detection algorithms on breast ultrasound images

Image segmentation is the partition of an image into a set of non-overlapping regions that comprise the entire image. The image is decomposed into meaningful parts, which are uniform with respect to certain characteristics, such as grey level or texture. This study presents a novel methodology to evaluate ultrasound image segmentation algorithms. The sonographic features can differentiate between various sized malignant and benign breast tumours. The clinical experiment can determine whether a tumour is benign or not, based on contour, shape, echogenicity and echo texture. Further study of the standardized sonographic features, especially the tumour contour and shape, will improve the positive predictive value and accuracy rate in breast tumour detection. The effectiveness of using this methodology is illustrated by evaluating image segmentation on breast ultrasound images. Via definite segmentation, the appreciated tumour shape and contour can be ascertained. Furthermore, this method can enhance the ability of ultrasound to distinguish between benign and malignant breast lesions.