Standalone computer-aided detection compared to radiologists' performance for the detection of mammographic masses.

OBJECTIVES: We developed a computer-aided detection (CAD) system aimed at decision support for detection of malignant masses and architectural distortions in mammograms. The effect of this system on radiologists' performance depends strongly on its standalone performance. The purpose of this study was to compare the standalone performance of this CAD system to that of radiologists. METHODS: In a retrospective study, nine certified screening radiologists and three residents read 200 digital screening mammograms without the use of CAD. Performances of the individual readers and of CAD were computed as the true-positive fraction (TPF) at a false-positive fraction of 0.05 and 0.2. Differences were analysed using an independent one-sample t-test. RESULTS: At a false-positive fraction of 0.05, the performance of CAD (TPF = 0.487) was similar to that of the certified screening radiologists (TPF = 0.518, P = 0.17). At a false-positive fraction of 0.2, CAD performance (TPF = 0.620) was significantly lower than the radiologist performance (TPF = 0.736, P <0.001). Compared to the residents, CAD performance was similar for all false-positive fractions. CONCLUSIONS: The sensitivity of CAD at a high specificity was comparable to that of human readers. These results show potential for CAD to be used as an independent reader in breast cancer screening.

Computer-aided detection of masses at mammography: interactive decision support versus prompts.

PURPOSE: To compare effectiveness of an interactive computer-aided detection (CAD) system, in which CAD marks and their associated suspiciousness scores remain hidden unless their location is queried by the reader, with the effect of traditional CAD prompts used in current clinical practice for the detection of malignant masses on full-field digital mammograms. MATERIALS AND METHODS: The requirement for institutional review board approval was waived for this retrospective observer study. Nine certified screening radiologists and three residents who were trained in breast imaging read 200 studies (63 studies containing at least one screen-detected mass, 17 false-negative studies, 20 false-positive studies, and 100 normal studies) twice, once with CAD prompts and once with interactive CAD. Localized findings were reported and scored by the readers. In the prompted mode, findings were recorded before and after activation of CAD. The partial area under the location receiver operating characteristic (ROC) curve for an interval of low false-positive fractions typical for screening, from 0 to 0.2, was computed for each reader and each mode. Differences in reader performance were analyzed by using software. RESULTS: The average partial area under the location ROC curve with unaided reading was 0.57, and it increased to 0.62 with interactive CAD, while it remained unaffected by prompts. The difference in reader performance for unaided reading versus interactive CAD was statistically significant (P = .009). CONCLUSION: When used as decision support, interactive use of CAD for malignant masses on mammograms may be more effective than the current use of CAD, which is aimed at the prevention of perceptual oversights.

Increase in perceived case suspiciousness due to local contrast optimisation in digital screening mammography.

OBJECTIVES: To determine the influence of local contrast optimisation on diagnostic accuracy and perceived suspiciousness of digital screening mammograms. METHODS: Data were collected from a screening region in the Netherlands and consisted of 263 digital screening cases (153 recalled,110 normal). Each case was available twice, once processed with a tissue equalisation (TE) algorithm and once with local contrast optimisation (PV). All cases had digitised previous mammograms. For both algorithms, the probability of malignancy of each finding was scored independently by six screening radiologists. Perceived case suspiciousness was defined as the highest probability of malignancy of all findings of a radiologist within a case. Differences in diagnostic accuracy of the processing algorithms were analysed by comparing the areas under the receiver operating characteristic curves (A(z)). Differences in perceived case suspiciousness were analysed using sign tests. RESULTS: There was no significant difference in A(z) (TE: 0.909, PV 0.917, P = 0.46). For all radiologists, perceived case suspiciousness using PV was higher than using TE more often than vice versa (ratio: 1.14-2.12). This was significant (P <0.0083) for four radiologists. CONCLUSIONS: Optimisation of local contrast by image processing may increase perceived case suspiciousness, while diagnostic accuracy may remain similar. KEY POINTS: Variations among different image processing algorithms for digital screening mammography are large. Current algorithms still aim for optimal local contrast with a low dynamic range. Although optimisation of contrast may increase sensitivity, diagnostic accuracy is probably unchanged. Increased local contrast may render both normal and abnormal structures more conspicuous.

Consequences of digital mammography in population-based breast cancer screening: initial changes and long-term impact on referral rates.

OBJECTIVES: To investigate the referral pattern after the transition to full-field digital mammography (FFDM) in a population-based breast cancer screening programme. METHODS: Preceding the nationwide digitalisation of the Dutch screening programme, an FFDM feasibility study was conducted. Detection and referral rates for FFDM and screen-film mammography (SFM) were compared for first and subsequent screens. Furthermore, radiological characteristics of referrals in digital screening were assessed. RESULTS: A total of 312,414 screening mammograms were performed (43,913 digital and 268,501 conventional), with 4,473 consecutive referrals (966 following FFDM). Initially the FFDM referral rate peaked, and many false-positive results were noted as a consequence of pseudolesions and increased detection of (benign) microcalcifications. A higher overall referral rate was observed in FFDM screening in both first and subsequent examinations (p < .001), with a significant increase in cancer detection (p = .010). CONCLUSION: As a result of initial inexperience with digital screening images implementing FFDM in a population-based breast cancer screening programme may lead to a strong, but temporary increase in referral. Dedicated training in digital screening for radiographers and screening radiologists is therefore recommended. Referral rates decrease and stabilise (learning curve effect) at a higher level than in conventional screening, yet with significantly enhanced cancer detection.

Breast cancer screening results 5 years after introduction of digital mammography in a population-based screening program.

PURPOSE: To compare full-field digital mammography (FFDM) using computer-aided diagnosis (CAD) with screen-film mammography (SFM) in a population-based breast cancer screening program for initial and subsequent screening examinations. MATERIALS AND METHODS: The study was approved by the regional medical ethics review board. Informed consent was not required. In a breast cancer screening facility, two of seven conventional mammography units were replaced with FFDM units. Digital mammograms were interpreted by using soft-copy reading with CAD. The same team of radiologists was involved in the double reading of FFDM and SFM images, with differences of opinion resolved in consensus. After 5 years, screening outcomes obtained with both modalities were compared for initial and subsequent screening examination findings. RESULTS: A total of 367,600 screening examinations were performed, of which 56,518 were digital. Breast cancer was detected in 1927 women (317 with FFDM). At initial screenings, the cancer detection rate was .77% with FFDM and .62% with SFM. At subsequent screenings, detection rates were .55% and .49%, respectively. Differences were not statistically significant. Recalls based on microcalcifications alone doubled with FFDM. A significant increase in the detection of ductal carcinoma in situ was found with FFDM (P < .01). The fraction of invasive cancers with microcalcifications as the only sign of malignancy increased significantly, from 8.1% to 15.8% (P < .001). Recall rates were significantly higher with FFDM in the initial round (4.4% vs 2.3%, P < .001) and in the subsequent round (1.7% vs 1.2%, P < .001). CONCLUSION: With the FFDM-CAD combination, detection performance is at least as good as that with SFM. The detection of ductal carcinoma in situ and microcalcification clusters improved with FFDM using CAD, while the recall rate increased.

Breast arterial calcifications are correlated with subsequent development of coronary artery calcifications, but their aetiology is predominantly different.

OBJECTIVE: To study whether calcifications in breast arteries, as seen on mammograms, predict future development of coronary artery calcifications. METHODS: We studied 499 women, aged 49-70 years, participating in a breast cancer screening program and investigated whether arterial calcifications in the breast (BAC) are associated with coronary arterial calcifications (CAC) after 9 years follow-up. Mammograms were reviewed for the presence of BAC. CAC was assessed by multi slice computed tomography (MSCT). With logistic regression analysis the independent effect of various risk factors on BAC and CAC was measured. RESULTS: BAC was present in 58 of 499 women (12%) and CAC score>0 was present in 262 of 499 women (53%). BAC was strongly associated with CAC (OR 3.2, 95% CI 1.71-6.04) and this remained significant after adjustment for age at baseline and the duration of follow-up (OR 2.1, 95% CI 1.10-4.23). Most CV risk factors were associated with CAC but not with BAC. Only parity was significantly associated with both increased CAC (OR 2.1, 95% CI 1.21-3.60) and increased BAC (OR 5.3, 95% CI 1.23-22.43). Breastfeeding was associated with BAC (OR 3.4, 95% CI 1.40-8.23) but not with CAC (OR 1.3, 95% CI 0.84-1.93). CONCLUSION: Breast arterial calcifications are predictive of subsequent development of calcifications in the coronary arteries.

Vitamin K intake and calcifications in breast arteries.

OBJECTIVES: Vitamin K is an important co-factor in the production of proteins that inhibit vascular calcification. A low dietary Vitamin K intake has been associated with aortic and coronary calcifications and an elevated cardiovascular risk. Calcifications in the arteries of the breasts have also been associated with cardiovascular risk, but whether there is a relation with a low Vitamin K intake has not yet been studied. METHODS: We conducted a cross-sectional study among 1689 women, aged 49-70 years. Dietary Vitamins K1 and K2 intake was calculated from a validated food frequency questionnaire. Breast arterial calcifications (BAC) were assessed on standard screening mammograms by two independent radiologists. With a general linear model mean Vitamins K1, K2 and Vitamin K2 subtypes were calculated for women with BAC and without, adjusted for age, smoking, diabetes, intake of saturated fat, mono-unsaturated fat, poly-unsaturated fat and protein- and calcium-intake. RESULTS: BAC was less common in the highest (9%) quartile of Vitamin K2 intake, compared to the lowest (13%) (OR 0.7, 95% CI 0.5-1.1) and not different across quartiles of Vitamin K1 intake. Mean Vitamin K2 levels and mean levels of Vitamin K2 subtypes MK-5 through MK-10 were lower in the participants with BAC (p=0.01) compared to participants without BAC. However, after adjustment for aging, smoking, diabetes and dietary factors the association of mean Vitamin K2 intake with BAC was no longer significant. CONCLUSION: Calcifications in breast arteries are not associated with a lower dietary intake of Vitamin K.

Importance of comparison of current and prior mammograms in breast cancer screening.

PURPOSE: To retrospectively determine the influence of comparing current mammograms with prior mammograms on breast cancer detection in screening and to investigate a protocol in which prior mammograms are viewed only when necessary. MATERIALS AND METHODS: Institutional review board approval was not required. Participants gave written informed consent. Twelve experienced screening radiologists read 160 soft-copy screening mammograms twice, once with and once without prior mammograms. Eighty mammograms were obtained in women in whom breast cancer was diagnosed later; the other 80 mammograms had been reported as normal or benign. All cancers were visible in retrospect. Readers located potential abnormalities, estimated likelihood of malignancy for each finding, and indicated whether prior mammograms were considered necessary. The effect of prior mammograms on detection was determined by computing the mean lesion localized fraction in a range of low fractions of nonlesion locations corresponding to operating points in screening. Scores for both reading sessions were combined to assess the effect of making prior mammograms available only when requested. Data were analyzed by comparing the number of localized lesions between the two reading conditions with a paired two-tailed Student t test and applying a linear mixed model to test differences in average mean lesion localized fraction between reading conditions. P values less than .05 indicated statistical significance. RESULTS: Without prior mammograms, significantly more annotations were made. When only positive cases were considered, no difference was observed. Reading performance was significantly better when prior screening mammograms were available. At fixed lesion localized fraction, nonlesion localized fraction was reduced by 44% (P<.001) on average when prior mammograms were read. Performance was also increased for combined reading mode (ie, when prior mammograms were available on request only). However, this increase was smaller than that when prior mammograms were always available. Prior mammograms were requested in 24%-33% of all cases and were requested more often in positive cases. CONCLUSION: Comparison with prior mammograms significantly improves overall performance and can reduce referrals due to nonlesion locations. Limiting the availability of prior mammograms to cases selected by the reader reduces the beneficial effect of prior mammograms.

Arterial calcifications seen on mammograms: cardiovascular risk factors, pregnancy, and lactation.

PURPOSE: To retrospectively assess if mammographic calcium deposits are related to coronary heart disease (CHD) risk factors and reproductive factors in a subset of women participating in the European Prospective Investigation into Cancer and Nutrition study. MATERIALS AND METHODS: The study was approved by the institutional review board of the University Medical Center Utrecht; informed consent was obtained. Mammograms were evaluated by two radiologists for the presence of breast arterial calcifications (BAC) in the Prospect cohort, a breast cancer screening population of women aged 49-70 years (mean, 57 years) within the European Prospective Investigation into Cancer and Nutrition study. Cardiovascular risk factors and reproductive factors were examined for independent effects on the prevalence of BAC. Logistic regression analysis was performed. RESULTS: BAC was present in 194 of 1699 (11%) women and increased with age to 20% in the highest quartile of age (mean, 66 years). The odds ratio was 4.7 in the highest versus the lowest quartile of age (95% confidence interval [CI], 2.9, 7.6). After adjustment for age, no significant association was found between BAC and traditional cardiovascular risk factors. Current smoking was inversely related to BAC (odds ratio, 0.6; 95% CI: 0.4, 0.9). BAC was prevalent in 2.5% of nulliparous women, in 9% of women with one or two children, and in 17% of women with three or more children (odds ratio, 7.2; 95% CI: 2.9, 18.0). Breast feeding after pregnancy was significantly associated with BAC in women who ever were pregnant (odds ratio, 2.2; 95% CI: 1.4, 3.6). CONCLUSION: Calcifications in arteries on mammograms are associated with increasing age, pregnancy, and lactation but not with various cardiovascular risk factors.