Search

Two graph theoretic methods are used in conjunction with active contours to segment the pectoral muscle in 82 screening mammograms. To validate the method, the boundaries are also marked by four radiologists with different levels of experience in mammography. The simultaneous truth and performance level estimation (STAPLE) method is used to estimate the true boundary and to estimate the sensitivity and specificity of the segmentation schemes. The performance of one of the two algorithms is found not differ significantly from radiologists. [ABSTRACT FROM AUTHOR]

Topological properties of the breast ductal network have shown the potential for classifying clinical breast images with and without radiological findings. In this paper, we review three methods for the description and classification of breast ductal topology. The methods are based on ramification matrices and symbolic representation via string encoding signatures. The performance of these methods has been compared using clinical x-ray and MR images of breast ductal networks. We observed the accuracy of the classification between the ductal trees segmented from the x-ray galactograms with radiological findings and normal cases in the range of 0.86-0.91%. The accuracy of the classification of the ductal trees segmented from the MR autogalactograms was observed in the range of 0.5-0.89%. [ABSTRACT FROM AUTHOR]

Mammographic risk assessment is concerned with the probability of a woman developing breast cancer. Recently, it has been suggested that the density of linear structures is related to risk. For 321 images from the MIAS database, a measure of line strength was obtained for each pixel using the Line Operator method. The proportion of pixels with line strength above a threshold level was calculated for each image and the results categorised by Tabar pattern, Boyd SCC class and BIRADS class. The results indicated a significant difference between Boyd classes 1-3 (low risk) and classes 4-6 (high risk), and between most Tabar patterns and BIRADS classes. [ABSTRACT FROM AUTHOR]

Several studies have showed that increased mammographic density is an important risk factor for breast cancer. Dense tissue often appears as textured regions in mammograms, so density and texture estimation are inextricably linked. It has been demonstrated that texture classes can be learned, and that subsequently textures can be classified using the joint distribution of intensity values over extremely compact neighbourhoods. Motivated by the success of texture classification, we propose an fully automated scheme for mammogram texture classification and segmentation. The classification method first has a training step to model the joint distribution for each breast density class. Subsequently, a statistical comparison is used to determine the class label for new images. Inspired by the classification, we combine the so-called image patch method with a HMRF(Hidden Markov Random Field) to achieve mammogram segmentation. [ABSTRACT FROM AUTHOR]

This paper describes an algorithm to segment mammo- graphic images into regions corresponding to different densities. The breast parenchymal segmentation uses information extracted for statistical texture based classification which is in turn incorporated in multi-vector Markov Random Fields. Such segmentation is key to developing quantitative mammographic analysis. The algorithm's performance is evaluated quantitatively and qualitatively and the results show the feasibility of segmenting different mammographic densities. [ABSTRACT FROM AUTHOR]

A method of extracting salient image features in mammograms at multiple scales using the monogenic signal is presented. The derived local phase provides structure information (such as edge, ridge etc.) while the local amplitude encodes the local brightness and contrast information. Together with the simultaneously computed orientation, these three pieces of information can be used for mammogram segmentation including locating the inner breast edge which is important for quantitative breast density assessment. Due to the contrast invariant property of the local phase, the algorithm proves to be very reliable on an extensive datasets of images obtained from various sources and digitized by different scanners. [ABSTRACT FROM AUTHOR]

We compared three conspicuity tests applied to four full field digital mammography (FFDM) systems. The tests included: 1) the calculation of noise equivalent quanta (NEQ); 2) contrast-detail analysis with the CDMAM 3.4 phantom and 3) evaluation of the detectability of (simulated) microcal-cifications with specific well-known dimensions in mastectomy images. For each contrast-resolution test method, the exposure, processing and viewing conditions were identical. As a result, the only variable for a given test was the physical performance of the detector. The three test methods each rank the detectors in the same order. The flat-panel detector ranked the best overall, the dual-sided read-out storage phosphor detector ranked second and the single-sided-read-out storage phosphor detectors with 50 μm and 100 μm pixel sizes ranked similarly and were inferior to the other 2 detectors. [ABSTRACT FROM AUTHOR]

Digital mammography is rapidly becoming a mature imaging modality. To maintain high quality in mammography, a routine quality control program is necessary to detect drifting or degradation of system performance over time. The American College of Radiology is developing a quality control program which will apply to all types of full-field digital mammography equipment, and provide effective and more efficient validation of performance. In the DMIST trial, there were no failures for many of the QC tests during the 24 months imaging was performed. When systems failed, they generally did so suddenly, rather than through gradual deterioration of performance. A recommended set of tests is presented, which can be used to ensure that full-field digital mammography (FFDM) systems are functioning correctly, and consistently producing mammograms of excellent image quality. Keywords: Digital mammography, quality control, image quality. [ABSTRACT FROM AUTHOR]

Purpose: To evaluate efficacy and diagnostic accuracy of BARCO LCD 5Mpixel displays, compared to BARCO CRT 5Mpixel displays in full-field-digital-mammography (FFDM) interpretation. Material and Methods: FFDM mammograms obtained by 100 patients, were analyzed by three independent radiologists experienced in breast imaging, using two different CRT and LCD displays. All cases were selected by a fourth radiologist in order to cover several possible ages and types of breast. Half of cases were negative and half were positive for malignancy, proven by percutaneous biopsy. Readers were blinded to history of patients, ultrasound examination and biopsy results. To minimize recall bias, an interval of at least 30 days between interpretations of each case on two different monitors was chosen. Each reader evaluated cases classifying them according the ACR BIRADS categories. Moreover, they assigned a rate (0-100) corresponding to the Probability of Malignancy (POM) of each case classified into BIRADS categories 3 to 5. Finally, they assigned a rate (0-100) corresponding to reading confidence. Analysis included ROC curves of POM for each doctor and for pooled data, sensitivity and specificity for the BIRADS≥3 and BIRADS≥4 thresholds for each doctor and for pooled data, and finally main results of "Multireader- Multicase ROC Analysis Of Variance". For each analysis a comparison was made between the two monitors. Results: No statistical significance was seen between the two displays regarding POM, sensitivity and specificity, nor for single reader either for pooled data. Conclusions: This study provides a reasonable assurance that the examined CRT and LCD display systems are comparable for FFDM interpretation. [ABSTRACT FROM AUTHOR]

We have developed computerized methods for the analysis of mammographic lesions in order to aid in the diagnosis of breast cancer. Our automatic methods include the extraction of the lesion from the breast parenchyma, the characterization of the lesion features in terms of mathematical descriptors, and the merging of these lesion features into an estimate of the probability of malignancy. Our initial development was performed on digitized screen film mammograms. We report our progress here in converting our methods for use with images from full-field digital mammography (FFDM). It is apparent from our initial comparisons on CAD for SFMD and FFDM that the overall concepts and image analysis techniques are similar, however reoptimization for a particular lesion segmentation or a particular mammographic imaging system are warranted. [ABSTRACT FROM AUTHOR]

The availability of the large dataset of screen/film and full-field digital mammograms acquired through the Digital Mammography Imaging Screening Trial (DMIST) presents an extraordinary opportunity for the assessment of CAD devices. The National Cancer Institute and the National Institute of Biomedical Imaging and Bioengineering at the U.S. National Institutes of Health have engaged FDA scientists in the development of a plan to leverage this imaging resource to benchmark the performance of current CAD systems. In this talk, we will present an initial proposal for utilizing the DMIST data to quantitatively assess current CAD systems. It is our goal to engage the IWDM community and other interested groups in the development of a consensus on acceptable study designs for this purpose. [ABSTRACT FROM AUTHOR]

Automated detection of masses on mammograms is challenged by the presence of dense breast parenchyma. The aim of this study is to investigate the feasibility of wavelet-based feature analysis in identifying spiculated and circumscribed masses in dense breast parenchyma. The method includes an edge detection step for breast border identification and employs Gaussian mixture modeling for dense parenchyma labeling. Subsequently, wavelet decomposition is performed and intensity as well as orientation features are extracted from approximation and detail subimages, respectively. Logistic regression analysis (LRA) is employed to differentiate spiculated and circumscribed masses from normal dense parenchyma. The proposed method is tested in 90 dense mammograms containing spiculated masses (30), circumscribed masses (30) and normal parenchyma (30). Free-response receiver operating characteristic (FROC) analysis is used to evaluate the performance of the method, achieving 83.3% sensitivity at 1.5 and 1.8 false positives per image for identifying spiculated and circumscribed masses, respectively. [ABSTRACT FROM AUTHOR]

A method that performs multiresolution enhancement, adaptive to breast components, for optimal visualization of the entire breast area is presented. The method includes an edge detection step to distinguish breast area from mammogram background and employs Gaussian mixture modeling to segment breast components (uncompressed fat, fat and dense). The original image is decomposed using a redundant discrete wavelet transform and magnitude coefficients corresponding to each breast component are linearly mapped for contrast enhancement. Coefficient mapping is controlled by a gain factor provided by the parameters of the modeled breast components. The processed image is derived by reconstruction of the modified wavelet coefficients. The algorithm is compared with two enhancement methods proposed for soft-copy display, in a dataset of 68 mammograms containing lesions. The proposed method demonstrates increased performance in accentuating lesions embedded in fatty or dense parenchyma, as well as in visualization of anatomical features in the entire breast area. [ABSTRACT FROM AUTHOR]

Digital breast tomosynthesis is becoming a clinically attractive modality based on its potential to combine the high resolution and high contrast images, and affordability of digital mammography, with the advantages of 3D image acquisition. In order to facilitate comparison of tomosynthesis images with previous mammographic exams of the same women, there is a need for a method of registering a mammogram with a tomosynthetic image of the same breast; this is the focus of this paper. We have chosen to approach this multimodality registration problem, starting from the simpler problem of registering a mammogram and the central tomosynthesis source image. Such a registration pair represents the most similar breast images obtained from different clinical modalities. In this study of 15 pairs of mammograms and central tomosynthesis projections of the same breast, on average we were able to compensate 94 percent of the per-pixel intensity differences that existed between the two images before the registration. [ABSTRACT FROM AUTHOR]

Breast tomosynthesis has the potential to improve lesion visibility and localization compared to conventional mammography. To be clinically useful, tomosynthesis must be able to achieve high image quality at acceptable radiation doses. Tomosynthesis data sets of simple low-contrast phantoms were acquired at varying dose levels. Image quality in the reconstructed volumes was analyzed by evaluating the voxel-to-voxel signal difference to noise ratio between a simulated lesion and the surrounding "tissue". Preliminary results indicate that image quality of small lesions is limited by scatter and reconstruction artifacts. In uniform backgrounds image quality appears to be quantum-noise limited, while in more complex backgrounds the structural noise tends to dominate. [ABSTRACT FROM AUTHOR]

Two observer experiments were performed to evaluate the performance of wavelet enhancement and compression methodologies for digitized mammography. One experiment was based on the localization response operating characteristic (LROC) model. The other estimated detection and localization accuracy rates. The results of both studies showed that the two algorithms consistently improved radiologists' performance although not always in a statistically significant way. An important outcome of this work was that lossy wavelet compression was as successful in improving the quality of digitized mammograms as the wavelet enhancement technique. The compression algorithm not only did not degrade the readers' performance but it improved it consistently while achieving compression rates in the range of 14 to 2051:1. The proposed wavelet algorithms yielded superior results for digitized mammography relative to conventional processing methodologies. Wavelets are valuable and diverse tools that could make digitized screen/film mammography equivalent to its direct digital counterpart leading to a filmless mammography clinic with full inter- and intra-system integration and real-time telemammography. [ABSTRACT FROM AUTHOR]

A method is proposed based on the combination of wavelet analysis and principal component analysis (PCA). Microcalcification (MC) candidate regions are initially labeled using area and contrast criteria. Mallat's redundant dyadic wavelet transform is used to analyze the frequency content of image patterns at horizontal and vertical directions. PCA is used to efficiently encode MC patterns in wavelet-decomposed images. Feature weights are computed from the projection of each candidate MC pattern at the wavelet-based principal components. To assess the effectiveness of the proposed method, the same analysis is carried out in original images. Candidate MC patterns are classified by means of Linear Discriminant Analysis (LDA). Free-response Receiver Operating Characteristic (FROC) curves are produced for identifying MC clusters. The highest performance is obtained when PCA is applied in wavelet decomposed images achieving 80% sensitivity at 0.5 false positives per image in a dataset with 50 subtle MC clusters in dense parenchyma. [ABSTRACT FROM AUTHOR]

Most of the CAD Systems for Mammograms are composed of algorithms analysing the four X-ray images individually. It is a general experience, that algorithms in search of microcalcification clusters can obtain high sensitivity only if specificity is low. To overcome efficiency problem this paper proposes a simple algorithm to combine information of the two views (MLO/CC) of the breast. The procedure is based upon the experiences of radiologists: masses and calcifications should emerge on both views, so if no matching is found, the given object is a false positive hit. A positioning system is evolved to find corresponding regions on the two images. Calcification clusters obtained in individual images are matched in "2.5-D" provided by the positioning system. The credibility value of the hit is reassessed by the matching. The proposed approach can significantly reduce the number of false positive hits in calcification. [ABSTRACT FROM AUTHOR]

The experience of clinical use of breast CAD system at the hospital of Gifu University School of Medicine was reported. The CAD system was Image Checker M1000-DM available for Senographe 2000D.During February 4, 2005 to May 16, 125 cases was examined by this device and 22cases of breast cancers were found. A case was misdiagnosed by radiologist before CAD, and CAD detected the lesion. Another case was correctly detected by radiologist before CAD, but CAD could not point out the lesion. 20 remain cases were detected correctly by both radiologist and CAD. CAD was supposed to be useful for the mammographic diagnosis of breast cancer. [ABSTRACT FROM AUTHOR]

The file size of images generated using digital mammography systems varies between 8 MB and 50 MB. The amount of data to be stored in digital screening programs is huge. Image compression may be helpful. In this study 8491 digital and digitised mammography images are compressed using 14 lossless compression schemes. The results show that using lossless image compression, the total amount of data to be stored can be reduced by a factor of 1.3 to 6.9 without loss of image quality. The actual data reduction depends strongly on the selected compression algorithm and the systems used to acquire and process the mammograms. The JPEG-LS and JPEG 2000 algorithms, both included in the DICOM standard, prove to be promising algorithms for screening programs because of the high compression ratios. [ABSTRACT FROM AUTHOR]

First 3D X-ray internal observation of DCIS (ductal carcinoma in-situ) is reported. Its rod shaped specimen with 3.6 mm in diameter and 4.7 mm in height was punched out to have successfully observed by using a newly made algorithm due to refraction for x-ray CT. Its data was acquired by the x-ray optics DEI (diffraction-enhanced imaging). Data of 900 projections with interval of 0.2 degrees was used at Photon Factory, KEK in Tsukuba. A reconstructed CT image may include clearly revealed ductus lactiferi, microcalcification and other structure. The voxel resolution is approximately 50 μm by the present instrumental condition. This modality could open up an x-ray pathological diagnosis. Keywords: X-ray refraction, X-ray dark-field imaging (XDFI), DEI, breast cancer, DCIS, pathological diagnosis, clinical diagnosis, ductus lactiferi. [ABSTRACT FROM AUTHOR]

In general, the use of more projections results in fewer tomosynthesis reconstruction artifacts. However, under a fixed dose, an excess number of projections will make the detector noise more pronounced in each of the x-ray shots and thus degrade image quality. Even in the absence of detector noise the advantages of higher projection numbers eventually have diminishing returns, making more projections unnecessary. In this study, we explore the dependence of tomosynthesis imaging performance on the number of projections, while keeping other factors fixed. We take the contrast-to-noise ratio as the figure of merit to search for the range of optimal projection number. The study is carried out through both simulations and experiments, with phantoms consisting of micro-calcification and mass objects, and a cadaver breast. The goal of this paper is to describe our methodology in general, and use a prototype tomosynthesis system as an example. The knowledge learned will help the design of future generation clinical tomosynthesis systems. [ABSTRACT FROM AUTHOR]

The purpose of this work was to investigate whether extra flat field corrections should be applied prior to the calculations of quality control quantities and whether there are necessary precautions regarding flat fielding in digital mammography. Effects from using one standard flat field correction for all imaging situations or absence of correction procedures were examined using homogenous PMMA slabs. Differences in field profiles for various exposure geometries and breast thicknesses were quantified. For three systems the maximal deviation of averaged pixel values (along a profile of ROIs parallel to the chest wall) varied from 1.3 % to 6.8 % over the whole image and from 0.6 % to 2.6 % if the analysis is limited to the central part. Extra flat field corrections are not necessary for most applications. If required, the corrections should be performed from images acquired with the same manual exposure and not after a time gap. [ABSTRACT FROM AUTHOR]

A new breast imaging approach is proposed and implemented that combines Magnetic Resonance Mammography and Electrical Impedance Scanning. In this paper, we report the results of a pilot study that demonstrated the feasibility of this new breast imaging approach. We also discuss our initial experience with the MR imaging parameters and sequences that are critical in observing the desired signal. [ABSTRACT FROM AUTHOR]