Computerized classification of benign and malignant masses on digitized mammograms: a study of robustness.

RATIONALE AND OBJECTIVES: The purpose of this study was to evaluate the robustness of a computerized method developed for the classification of benign and malignant masses with respect to variations in both case mix and film digitization. MATERIALS AND METHODS: The classification method included automated segmentation of mass regions, automated feature-extraction, and automated lesion characterization. The method was evaluated independently with a 110-case database consisting of 50 malignant and 60 benign cases. Mammograms were digitized twice with two different digitizers (Konica and Lumisys). Performance of the method in differentiating benign from malignant masses was evaluated with receiver operating characteristic (ROC) analysis. Effects of variations in both case mix and film digitization on performance of the method also were assessed. RESULTS: Categorization of lesions as malignant or benign with an artificial neural network (or a hybrid) classifier achieved an area under the ROC curve, Az, value of 0.90 (0.94 for the hybrid) on the previous training database in a round-robin evaluation and Az values of 0.82 (0.81) and 0.81 (0.82) on the independent database for the Konica and Lumisys formats, respectively. These differences, however, were not statistically significant (P > .10). CONCLUSION: The computerized method for the classification of lesions on mammograms was robust with respect to variations in case mix and film digitization.

Computerized analysis of mammographic parenchymal patterns for breast cancer risk assessment: feature selection.

Our purpose in this study was to identify computer-extracted, mammographic parenchymal patterns that are associated with breast cancer risk. We extracted 14 features from the central breast region on digitized mammograms to characterize the mammographic parenchymal patterns of women at different risk levels. Two different approaches were employed to relate these mammographic features to breast cancer risk. In one approach, the features were used to distinguish mammographic patterns seen in low-risk women from those who inherited a mutated form of the BRCA1/BRCA2 gene, which confers a very high risk of developing breast cancer. In another approach, the features were related to risk as determined from existing clinical models (Gail and Claus models), which use well-known epidemiological factors such as a woman's age, her family history of breast cancer, reproductive history, etc. Stepwise linear discriminant analysis was employed to identify features that were useful in differentiating between "low-risk" women and BRCA1/BRCA2-mutation carriers. Stepwise linear regression analysis was employed to identify useful features in predicting the risk, as estimated from the Gail and Claus models. Similar computer-extracted mammographic features were identified in the two approaches. Results show that women at high risk tend to have dense breasts and their mammographic patterns tend to be coarse and low in contrast.

Computerized analysis of lesions in US images of the breast.

RATIONALE AND OBJECTIVES: Breast sonography is not routinely used to distinguish benign from malignant solid masses because of considerable overlap in their sonographic appearances. The purpose of this study was to investigate the computerized analyses of breast lesions in ultrasonographic (US) images in order to ultimately aid in the task of discriminating between malignant and benign lesions. MATERIALS AND METHODS: Features related to lesion margin, shape, homogeneity (texture), and posterior acoustic attenuation pattern in US images of the breast were extracted and calculated. The study database contained 184 digitized US images from 58 patients with 78 lesions. Benign lesions were confirmed at biopsy or cyst aspiration or with image interpretation alone; malignant lesions were confirmed at biopsy. Performance of the various individual features and output from linear discriminant analysis in distinguishing benign from malignant lesions was studied by using receiver operating characteristic (ROC) analysis. RESULTS: At ROC analysis, the feature characterizing the margin yielded Az values (area under the ROC curve) of 0.85 and 0.75 in distinguishing between benign and malignant lesions for the entire database and for an "equivocal" database, respectively. The equivocal database contained lesions that had been proved to be benign or malignant at cyst aspiration or biopsy. Linear discriminant analysis round-robin runs yielded Az values of 0.94 and 0.87 in distinguishing benign from malignant lesions for the entire database and for the equivocal database, respectively. CONCLUSION: Computerized analysis of US images has the potential to increase the specificity of breast sonography.

Automated computerized classification of malignant and benign masses on digitized mammograms.

RATIONALE AND OBJECTIVES: To develop a method for differentiating malignant from benign masses in which a computer automatically extracts lesion features and merges them into an estimated likelihood of malignancy. MATERIALS AND METHODS: Ninety-five mammograms depicting masses in 65 patients were digitized. Various features related to the margin and density of each mass were extracted automatically from the neighborhoods of the computer-identified mass regions. Selected features were merged into an estimated likelihood of malignancy by, using three different automated classifiers. The performance of the three classifiers in distinguishing between benign and malignant masses was evaluated by receiver operating characteristic analysis and compared with the performance of an experienced mammographer and that of five less experienced mammographers. RESULTS: Our computer classification scheme yielded an area under the receiver operating characteristic curve (Az) value of 0.94, which was similar to that for an experienced mammographer (Az = 0.91) and was statistically significantly higher than the average performance of the radiologists with less mammographic experience (Az = 0.81) (P = .013). With the database used, the computer scheme achieved, at 100% sensitivity, a positive predictive value of 83%, which was 12% higher than that for the performance of the experienced mammographer and 21% higher than that for the average performance of the less experienced mammographers (P < .0001). CONCLUSION: Automated computerized classification schemes may be useful in helping radiologists distinguish between benign and malignant masses and thus reducing the number of unnecessary biopsies.

Clinical outcome of doubtful mammographic findings.

OBJECTIVE: We performed this study to determine whether subthreshold findings of malignancy can be detected prospectively during interpretation of routine mammographic screenings. MATERIALS AND METHODS: For several years we have marked with crayon mammographic findings identified at screening that may represent precursor lesions but that we judge so likely to be benign that further workup is not indicated. The crayon marks prompt a closer look at the designated area on the next screening examination. We reviewed all screening examinations performed between August 1991 and October 1992 for which we had done prior normal screening. The prior mammograms were examined for crayon marks that indicated possible precursor lesions, and outcome was assessed on the basis of findings of subsequent screenings. RESULTS: Crayon marks of 543 findings (382 women) were identified on prior normal examinations in the 5514 cases reviewed. Marked findings consisted of calcifications (48%), noncalcified nodules (22%), vague densities (18%), asymmetries (7%), combination findings (2%), and possible architectural distortions (2%). On subsequent examination (mean interval, 30 months; range, 231 days to 90 months), 74% of marked findings showed no significant change, 21% were less apparent or had disappeared, 4% were slightly more apparent but still within normal limits, and six cases (1%) were interpreted as abnormal, which resulted in diagnostic workup. Three of the six latter cases were judged to be benign after further imaging evaluation; the other three patients underwent biopsy. One of these cases resulted in a diagnosis of low-grade ductal carcinoma in situ; the other two findings were benign. No unsuspected cancers were found after linkage with our tumor registry 1 year after the most recent examination with crayon marks. CONCLUSION: This study shows that almost all prospectively marked benign-appearing findings are benign and inconsequential. The medicolegal implications of this observation include the inappropriateness of malpractice claims in cases in which cancer could not reasonably be detected prospectively.

Malignant and benign clustered microcalcifications: automated feature analysis and classification.

PURPOSE: To develop a method for differentiating malignant from benign clustered microcalcifications in which image features are both extracted and analyzed by a computer. MATERIALS AND METHODS: One hundred mammograms from 53 patients who had undergone biopsy for suspicious clustered microcalcifications were analyzed by a computer. Eight computer-extracted features of clustered microcalcifications were merged by an artificial neural network. Human input was limited to initial identification of the microcalcifications. RESULTS: Computer analysis allowed identification of 100% of the patients with breast cancer and 82% of the patients with benign conditions. The accuracy of computer analysis was statistically significantly better than that of five radiologists (P = .03). CONCLUSION: Quantitative features can be extracted and analyzed by a computer to distinguish malignant from benign clustered microcalcifications. This technique may help radiologists reduce the number of false-positive biopsy findings.

Analysis of spiculation in the computerized classification of mammographic masses.

Spiculation is a primary sign of malignancy for masses detected by mammography. In this study, we developed a technique that analyzes patterns and quantifies the degree of spiculation present. Our current approach involves (1) automatic lesion extraction using region growing and (2) feature extraction using radial edge-gradient analysis. Two spiculation measures are obtained from an analysis of radial edge gradients. These measures are evaluated in four different neighborhoods about the extracted mammographic mass. The performance of each of the two measures of spiculation was tested on a database of 95 mammographic masses using ROC analysis that evaluates their individual ability to determine the likelihood of malignancy of a mass. The dependence of the performance of these measures on the choice of neighborhood was analyzed. We have found that it is only necessary to accurately extract an approximate outline of a mass lesion for the purposes of this analysis since the choice of a neighborhood that accommodates the thin spicules at the margin allows for the assessment of margin spiculation with the radial edge-gradient analysis technique. The two measures performed at their highest level when the surrounding periphery of the extracted region is used for feature extraction, yielding Az values of 0.83 and 0.85, respectively, for the determination of malignancy. These are similar to that achieved when a radiologist's ratings of spiculation (Az = 0.85) are used alone. The maximum value of one of the two spiculation measures (FWHM) from the four neighborhoods yielded an Az of 0.88 in the classification of mammographic mass lesions.

Automated segmentation of digitized mammograms.

RATIONALE AND OBJECTIVES: Fast and reliable segmentation of digital mammograms into breast and nonbreast regions is an important prerequisite for further image analysis. We are developing a segmentation algorithm that is fully automated and can operate independent of type of digitizing system, image orientation, and image projection. METHODS: The algorithm identifies unexposed and direct-exposure image regions and generates a border surrounding the valid breast region, which can then be used as input for further image analysis. The program was tested on 740 digitized mammograms; the segmentation results were evaluated by two expert mammographers and two medical physicists. RESULTS: In 97% of the mammograms, the segmentation results were rated as acceptable for use in computer-aided diagnostic schemes. Segmentation problems encountered in the remaining 22 images (2.9%) were most often caused by digitization artifacts or poor mammographic technique. CONCLUSION: The developed algorithm can serve as a component of an "intelligent" workstation for computer-aided diagnosis in mammography.