The purpose of this paper is to present a new segmentation routine developed for mammographic masses. We previously developed a computer-aided detection (CAD) system for mammographic masses that employed a simple but imprecise segmentation procedure. To improve the systems performance, an iterative, linear segmentation routine was developed. The routine begins by employing a linear discriminant function to determine the optimal threshold between estimates of an objects interior and exterior pixels. After applying the threshold and identifying the objects outline, two constraints are applied to minimize the influence of extraneous background structures. Each iteration further refines the outline until the stopping criterion is reached. The segmentation algorithm was tested on a database of 181 mammographic images that contained forty-nine malignant and fifty benign masses. A set of suspicious regions of interest (ROIs) was found using the previous CAD system. Twenty features were measured from the regions before and after applying the new segmentation routine. The difference in the features discriminatory ability was examined via receiver operating characteristic (ROC) analysis. A significant performance difference was observed in many features, particularly those describing the object border. Free-response ROC (FROC) curves were utilized to examine how the overall CAD system performance changed with the inclusion of the segmentation routine. The FROC performance appeared to be improved, especially for malignant masses. When detecting 90% of the malignant masses, the previous system achieved 4.4 false positives per image (FPpI) compared to the post-segmentation systems 3.7 FPpI. At 85%, the respective FPpI are 4.1 and 2.1.
Bi-plane correlation imaging (BCI) is a new imaging approach that utilizes angular information from a bi-plane digital acquisition in conjunction with computer assisted detection (CAD) to reduce the degrading influence of anatomical noise in the detection of subtle lesions in planar images. An anthropomorphic chest phantom, supplemented with added nodule phantoms (5-13 mm at the image plane), was imaged from different posterior projections within a ±12° range by moving the x-ray tube vertically and horizontally with respect to the detector. Each image was analyzed using a basic front-end single-view CAD algorithm. The correlation of the suspect lesions from the PA view with those from each of the oblique views was examined using a priori knowledge of the acquisition geometry. The correlated suspect lesions were registered as positive. Using an optimum --3° vertical geometry and processing parameters, BCI resulted in 62.5% sensitivity, 1.5 FP/image, and 0.885 PPV. The corresponding values from the observer experiment were 56% sensitivity, 10.8 FP/image, and 0.45 PPV, respectively. Compared to single-view CAD results, the BCI reduced sensitivity by 20%. However, the corresponding reduction in FPs was notably higher (94%) leading to 140% improvement in the PPV. Changes in processing parameters could result in higher PPV and lower FP/image at the expense of lower sensitivity. Similar findings were indicated for small (5-9 mm) and large (9-13 mm) nodules, but the relative improvement was significantly higher for smaller nodules. (The research was supported by a grant from the NIH, R21CA91806.)
KEYWORDS: Databases, Mammography, CAD systems, Computer aided diagnosis and therapy, Cancer, Data modeling, Medical imaging, Solid modeling, Statistical analysis, Breast cancer
We propose to investigate the use of a Laguerre-Gauss Channelized Hotelling Observer (LG-CHO) for the basis of a computer aided detection scheme for masses in mammography. A database of 1320 regions of interest was selected from the DDSM database collected by the University of South Florida. The breakdown of the cases was: 656 normals, 307 benigns, and 357 cancers. For the detection task, cancer and benign cases were considered positive and normal was considered negative. A 25 channel LG-CHO was designed to best classify regions as containing a mass or not. Application of this LG-CHO to the database gave a ROC area under the curve of 0.936 and a partial area of 0.648. Additionally, at 98% sensitivity the classifier had a specificity of 44.8% and a positive predictive value of 64.2%. Preliminary results suggest that using a LG-CHO can provide a strong backbone for a CAD scheme to help radiologists with detection. These initial results should be able to be incorporated into a larger CAD system for higher performance either as a false positive reduction scheme or as an initial filter used for mass detection.
In this paper, we present preliminary results from a highly sensitive and specific CAD system for mammographic masses. For false positive reduction, the system incorporated features derived from shape, fractal, and channelized Hotelling observer (CHO) measurements. The database for this study consisted of 80 craniocaudal mammograms randomly extracted from USF's digital database for screening mammography. The database contained 49 mass findings (24 malignant, 25 benign). To detect initial mass candidates, a difference of Gaussians (DOG) filter was applied through normalized cross correlation. Suspicious regions were localized in the filtered images via multi-level thresholding. Features extracted from the regions included shape, fractal dimension, and the output from a Laguerre-Gauss (LG) CHO. Influential features were identified via feature selection techniques. The regions were classified with a linear classifier using leave-one-out training/testing. The DOG filter achieved a sensitivity of 88% (23/24 malignant, 20/25 benign). Using the selected features, the false positives per image dropped from ~20 to ~5 with no loss in sensitivity. This preliminary investigation of combining multi-level thresholded DOG-filtered images with shape, fractal, and LG-CHO features shows great promise as a mass detector. Future work will include the addition of more texture and mass-boundary descriptive features as well as further exploration of the LG-CHO.
We describe a probit regression approach for maximum-likelihood (ML) estimation of a linear observer template from human-observer data in two-alternative forced-choice experiments. Like a previous approach to ML estimation in this problem [Abbey & Eckstein, Proc. SPIE, Vol. 4324, 2001], our approach does not make any assumptions about the distribution of the images. The previous approach utilized a regularizing prior distribution to control the degrees of freedom in the problem. In this work, we constrain the observer template to be represented by a limited number of linear features. Standard methods of probit regression are described for estimating the feature weights, and hence the observer templates. We have used this probit regression method to estimate human-observer templates for the detection of a small (5mm diameter) round simulated mass embedded in digitized mammograms. Our estimated templates for detecting the mass contain a band of heavily weighted spatial frequencies from 0.08 to 0.3 cycles/mm. We show comparisons between the human-observer template data, and the templates of a number of linear model observers that have been investigated as perceptual models of the human.
This paper describes the development of a computer-aided diagnosis (CAD) tool for solitary pulmonary nodules. This CAD tool is built upon physically meaningful features that were selected because of their relevance to shape and texture. These features included a modified version of the Hotelling statistic (HS), a channelized HS, three measures of fractal properties, two measures of spicularity, and three manually measured shape features. These features were measured from a difficult database consisting of 237 regions of interest (ROIs) extracted from digitized chest radiographs. The center of each 256x256 pixel ROI contained a suspicious lesion which was sent to follow-up by a radiologist and whose nature was later clinically determined. Linear discriminant analysis (LDA) was used to search the feature space via sequential forward search using percentage correct as the performance metric. An optimized feature subset, selected for the highest accuracy, was then fed into a three layer artificial neural network (ANN). The ANN's performance was assessed by receiver operating characteristic (ROC) analysis. A leave-one-out testing/training methodology was employed for the ROC analysis. The performance of this system is competitive with that of three radiologists on the same database.
We propose to investigate a novel use of the Hotelling observer for the task of discrimination of solitary pulmonary nodules from a database of regions that were all deemed suspicious. A database of 239 regions of interest (ROIs) was collected from digitized chest radiographs. Each of these 256x256 pixel ROIs contained a suspicious lesion in the center for which we have a truth file. For our study, 25 separate Hotelling observers were set up in a 5x5 grid across the center of the ROIs. Each separate observer was designed to 'observe' a 15x15 pixel area of the image. Leave-one-out training was used to generate 25 output observer features. These 25 features were then narrowed down using a sequential forward searching linear discriminant analysis. The forward search was continued until the accuracy declined at 13 features and the subset was used as the input layer to an artificial neural network (ANN). This network was trained to minimize mean squared error and the output was the area under the ROC curve. The trained ANN gave an ROC area of .86. In comparison, three radiologists performed at ROC area indexes of .72, .79, and .83.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.