Texture features from mammographic images and risk of breast cancer.

Mammographic percent density (PD) is a strong risk factor for breast cancer, but there has been relatively little systematic evaluation of other features in mammographic images that might additionally predict breast cancer risk. We evaluated the association of a large number of image texture features with risk of breast cancer using a clinic-based case-control study of digitized film mammograms, all with screening mammograms before breast cancer diagnosis. The sample was split into training (123 cases and 258 controls) and validation (123 cases and 264 controls) data sets. Age-adjusted and body mass index (BMI)-adjusted odds ratios (OR) per SD change in the feature, 95% confidence intervals, and the area under the receiver operator characteristic curve (AUC) were obtained using logistic regression. A bootstrap approach was used to identify the strongest features in the training data set, and results for features that validated in the second half of the sample were reported using the full data set. The mean age at mammography was 64.0+/-10.2 years, and the mean time from mammography to breast cancer was 3.7+/-1.0 (range, 2.0-5.9 years). PD was associated with breast cancer risk (OR, 1.49; 95% confidence interval, 1.25-1.78). The strongest features that validated from each of several classes (Markovian, run length, Laws, wavelet, and Fourier) showed similar ORs as PD and predicted breast cancer at a similar magnitude (AUC=0.58-0.60) as PD (AUC=0.58). All of these features were automatically calculated (unlike PD) and measure texture at a coarse scale. These features were moderately correlated with PD (r=0.39-0.76), and after adjustment for PD, each of the features attenuated only slightly and retained statistical significance. However, simultaneous inclusion of these features in a model with PD did not significantly improve the ability to predict breast cancer.

An automated approach for estimation of breast density.

Breast density is a strong risk factor for breast cancer; however, no standard assessment method exists. An automated breast density method was modified and compared with a semi-automated, user-assisted thresholding method (Cumulus method) and the Breast Imaging Reporting and Data System four-category tissue composition measure for their ability to predict future breast cancer risk. The three estimation methods were evaluated in a matched breast cancer case-control (n = 372 and n = 713, respectively) study at the Mayo Clinic using digitized film mammograms. Mammograms from the craniocaudal view of the noncancerous breast were acquired on average 7 years before diagnosis. Two controls with no previous history of breast cancer from the screening practice were matched to each case on age, number of previous screening mammograms, final screening exam date, menopausal status at this date, interval between earliest and latest available mammograms, and residence. Both Pearson linear correlation (R) and Spearman rank correlation (r) coefficients were used for comparing the three methods as appropriate. Conditional logistic regression was used to estimate the risk for breast cancer (odds ratios and 95% confidence intervals) associated with the quartiles of percent breast density (automated breast density method, Cumulus method) or Breast Imaging Reporting and Data System categories. The area under the receiver operator characteristic curve was estimated and used to compare the discriminatory capabilities of each approach. The continuous measures (automated breast density method and Cumulus method) were highly correlated with each other (R = 0.70) but less with Breast Imaging Reporting and Data System (r = 0.49 for automated breast density method and r = 0.57 for Cumulus method). Risk estimates associated with the lowest to highest quartiles of automated breast density method were greater in magnitude [odds ratios: 1.0 (reference), 2.3, 3.0, 5.2; P trend < 0.001] than the corresponding quartiles for the Cumulus method [odds ratios: 1.0 (reference), 1.7, 2.1, and 3.8; P trend < 0.001] and Breast Imaging Reporting and Data System [odds ratios: 1.0 (reference), 1.6, 1.5, 2.6; P trend < 0.001] method. However, all methods similarly discriminated between case and control status; areas under the receiver operator characteristic curve were 0.64, 0.63, and 0.61 for automated breast density method, Cumulus method, and Breast Imaging Reporting and Data System, respectively. The automated breast density method is a viable option for quantitatively assessing breast density from digitized film mammograms.

Association of mammographic density with the pathology of subsequent breast cancer among postmenopausal women.

BACKGROUND: Limited studies have examined the associations between mammographic density and subsequent breast tumor characteristics. METHODS: Eligible women were part of a case-control study of postmenopausal breast cancer, were 40 years or older and had a routine mammogram 4 years or more before their diagnosis. Mammographic density (percent density, dense area, and nondense area) was estimated using a computer-assisted thresholding program. At the time of cancer diagnosis, cases were classified as asymptomatic or symptomatic based on medical record review and breast imaging workup. Pathologic review was done blinded to the density status. Linear regression models and tests for trend examined the association between pathologic characteristics of the breast tumor and the components of density for all participants, and stratified by symptom status at diagnosis. RESULTS: Of the 286 eligible cases, 77% were 60 years or older and mean percent density was 29.5% (SD, 14.6%). Density was not significantly associated with tumor size (P = 0.22), histologic type (P = 0.77), estrogen receptor (P = 0.11) or progesterone receptor (P = 0.37) status, mitotic activity (P = 0.12), or nuclear pleomorphism (P = 0.09; P values for percent density). An inverse association was suggested between tumor grade and percent density (32.0%, 30.3%, 26.7% for grades 1-3; P = 0.06 for trend). The inverse association with tumor grade and its components (nuclear pleomorphism and tubular differentiation) was only evident among the 97 symptomatic women; positive associations of estrogen receptor (P = 0.009) and progesterone receptor (P = 0.04) were also seen with percent density only in this subgroup. CONCLUSIONS: The inverse association between tumor grade and percent density in the symptomatic population could inform the biology of the association between mammographic density and breast cancer risk.

Noncathartic CT colonography with stool tagging: performance with and without electronic stool subtraction.

OBJECTIVE: The purpose of our study was to evaluate the performance of noncathartic, dietary unrestricted CT colonography, without and with the aid of electronic stool subtraction, for detecting colorectal neoplasia in a high-prevalence referral population. MATERIALS AND METHODS: Patients with known or suspected colorectal neoplasms were potentially eligible for participation, regardless of the presence or absence of gastrointestinal symptoms. Subjects ingested 21.6 g of barium in nine divided doses. CT colonography was performed in the standard fashion. Data sets were randomly evaluated by two of three experienced radiologists, with subsequent reanalysis of each data set after electronic stool subtraction at least 6 weeks later. Optical colonoscopy was performed after purgation and served as the reference standard. RESULTS: One hundred thirty-one adenomatous neoplasms were identified among 114 subjects. On a per subject basis, the sensitivity for detecting adenomas 6-9 or > or = 10 mm in diameter ranged from 53% to 88% and 84% to 93% without stool subtraction, respectively. By including stool subtraction, these sensitivity estimates improved to 68% to 92% and 93% to 94%, respectively. Specificity ranged from 71% to 91% and 88% to 100% for lesions 6-9 and > or = 10 mm in size, respectively. Double reading resulted in detection of 27 (87%) of 31 and 65 (96%) of 68 patients with 6-9 and > or = 10 mm adenomas, respectively. With double reading, the area under the receiver operating characteristic curve for large adenomas was 0.97. CONCLUSION: In this increased-risk referral population, CT colonography in the non-cathartic-tagged colon without dietary restrictions compared favorably with optical colonoscopy.

Longitudinal trends in mammographic percent density and breast cancer risk.

BACKGROUND: Mammographic density is a strong risk factor for breast cancer. However, whether changes in mammographic density are associated with risk remains unclear. MATERIALS AND METHODS: A study of 372 incident breast cancer cases and 713 matched controls was conducted within the Mayo Clinic mammography screening practice. Controls were matched on age, exam date, residence, menopause, interval between, and number of mammograms. All serial craniocaudal mammograms 10 years before ascertainment were digitized, and quantitative measures of percent density (PD) were estimated using a thresholding method. Data on potential confounders were abstracted from medical records. Logistic regression models with generalized estimating equations were used to evaluate the interactions among PD at earliest mammogram, time from earliest to each serial mammogram, and absolute change in PD between the earliest and subsequent mammograms. Analyses were done separately for PD measures from the ipsilateral and contralateral breast and also by use of hormone therapy (HT). RESULTS: Subjects had an average of five mammograms available, were primarily postmenopausal (83%), and averaged 61 years at the earliest mammogram. Mean PD at earliest mammogram was higher for cases (31%) than controls (27%; ipsilateral side). There was no evidence of an association between change in PD and breast cancer risk by time. Compared with no change, an overall reduction of 10% PD (lowest quartile of change) was associated with an odds ratio of 0.9997 and an increase of 6.5% PD (highest quartile of change) with an odds ratio of 1.002. The same results held within the group of 220 cases and 340 controls never using HT. Among the 124 cases and 337 controls known to use HT during the interval, there was a statistically significant interaction between change in PD and time since the earliest mammogram (P = 0.01). However, in all groups, the risk associated with the earliest PD remained a stronger predictor of risk than change in PD. CONCLUSION: We observed no association between change in PD with breast cancer risk among all women and those never using HT. However, the interaction between change in PD and time should be evaluated in other populations.

Mammographic breast density as a general marker of breast cancer risk.

Mammographic breast density is a strong risk factor for breast cancer but whether breast density is a general marker of susceptibility or is specific to the location of the eventual cancer is unknown. A study of 372 incident breast cancer cases and 713 matched controls was conducted within the Mayo Clinic mammography screening practice. Mammograms on average 7 years before breast cancer were digitized, and quantitative measures of percentage density and dense area from each side and view were estimated. A regional density estimate accounting for overall percentage density was calculated from both mammogram views. Location of breast cancer and potential confounders were abstracted from medical records. Conditional logistic regression was used to estimate associations, and C-statistics were used to evaluate the strength of risk prediction. There were increasing trends in breast cancer risk with increasing quartiles of percentage density and dense area, irrespective of the side of the breast with cancer (P(trends) < 0.001). Percentage density from the ipsilateral side [craniocaudal (CC): odds ratios (ORs), 1.0 (ref), 1.7, 3.1, and 3.1; mediolateral oblique (MLO): ORs, 1.0 (ref), 1.5, 2.2, and 2.8] and the contralateral side [CC: ORs, 1.0 (ref), 1.8, 2.2, and 3.7; MLO: ORs, 1.0 (ref), 1.6, 1.9, and 2.5] similarly predicted case-control status (C-statistics, 0.64-65). Accounting for overall percentage density, density in the region where the cancer subsequently developed was not a significant risk factor [CC: 1.0 (ref), 1.3, 1.0, and 1.2; MLO: 1.0 (ref), 1.1, 1.0, and 1.1 for increasing quartiles]. Results did not change when examining mammograms 3 years on average before the cancer. Overall mammographic density seems to represent a general marker of breast cancer risk that is not specific to breast side or location of the eventual cancer.

Development of a cathartic-free colorectal cancer screening test using virtual colonoscopy: a feasibility study.

OBJECTIVE: The purpose of our study was to develop a method to subtract barium-labeled stool from the colon using a phantom and to evaluate the performance of the technique in a pilot human population. MATERIALS AND METHODS: A phantom containing 6-mm flat polyps and three types of simulated stool (homogeneous, moderately heterogeneous, and severely heterogeneous) mixed with barium was created, scanned, and tested using three stool subtraction algorithms but no cathartic. Thirty patients with suspected colorectal polyps were studied using stool tagging to determine which was the most effective stool subtraction algorithm. Colonoscopy was the reference standard. Examinations were evaluated blindly using the unsubtracted and 6 weeks later both the unsubtracted and subtracted data sets. RESULTS: A threshold of 200 H and expansion and convolution techniques were the most effective tools for subtracting stool and minimizing artifacts. When applied to the human population, sensitivities using the unsubtracted data sets were 90% (18/20) and 68% (26/38) for polyps > or = 1 cm and > or = 5 mm, respectively. Specificities were 100% (4/4) and 75% (3/4) for polyps > or = 1 cm and > or = 5 mm. For the stool-subtracted data sets, sensitivities were 90% (18/20) and 71% (27/38) for polyps > or = 1 cm and > or = 5 mm. Per patient sensitivities were 88% (15/17) and 77% (20/26) for > or = 1 cm and > or = 5 mm polyps. Specificities were 100% (4/4) for large polyps and 25% (1/4) for smaller polyps. CONCLUSION: Image processing tools combining thresholding, expansion, and convolution were the most useful for stool subtraction. Laxative-free colon examinations using barium for stool labeling can be performed at CT colonography with or without stool subtraction with high accuracy. Further study is warranted.

Visceral fat analysis at CT colonography.

RATIONALE AND OBJECTIVES: Obesity is associated with increased risks for colorectal neoplasia. Few studies have examined quantitative body fat measurements as predictors of colorectal polyps. The objective is to determine whether visceral fat is associated with colorectal polyps at computed tomography (CT) colonography. MATERIALS AND METHODS: Case (n = 25) and control (n = 25) subjects with proven large (>1 cm) colorectal adenomas or normal colons respectively were randomly selected from among an established CT colonography research study cohort. Using supine CT colonography data, the body wall was traced at three levels: top of the right kidney, iliac crest, and superior acetabulum. Total area from the three slices and each slice area were determined within the visceral fat range (-170 to -45 Hounsfield units) and recorded within the selected region. Visceral fat measures were compared between patient groups with and without polyps. RESULTS: None of the single slice visceral fat area measures or summed measures predicted case or control status. The most informative visceral fat measure was obtained at the top of the right kidney with a maximum area under the received operator characteristic curve of 0.77 (0.05 SE). For a selected sensitivity of 75%, the maximum specificity for a large (>or=1 cm) polyp was 64%. CONCLUSION: In this pilot study, visceral fat measures at CT colonography were not significantly associated with the presence of large colorectal adenomas. However, odd ratios were elevated by a factor of 2. This suggests that a larger study may be justified.