Search

BACKGROUND AND OBJECTIVE: Breast density assessed from digital mammograms is a biomarker for higher risk of developing breast cancer. Experienced radiologists assess breast density using the Breast Image and Data System (BI-RADS) categories. Supervised learning algorithms have been developed with this objective in mind, however, the performance of these algorithms depends on the quality of the ground-truth information which is usually labeled by expert readers. These labels are noisy approximations of the ground truth, as there is often intra- and inter-reader variability among labels. Thus, it is crucial to provide a reliable method to obtain digital mammograms matching BI-RADS categories. This paper presents RegL (Labels Regularizer), a methodology that includes different image pre-processes to allow both a correct breast segmentation and the enhancement of image quality through an intensity adjustment, thus allowing the use of deep learning to classify the mammograms into BI-RADS categories. The Confusion Matrix (CM) - CNN network used implements an architecture that models each radiologist's noisy label. The final methodology pipeline was determined after comparing the performance of image pre-processes combined with different DL architectures. METHODS: A multi-center study composed of 1395 women whose mammograms were classified into the four BI-RADS categories by three experienced radiologists is presented. A total of 892 mammograms were used as the training corpus, 224 formed the validation corpus, and 279 the test corpus. RESULTS: The combination of five networks implementing the RegL methodology achieved the best results among all the models in the test set. The ensemble model obtained an accuracy of (0.85) and a kappa index of 0.71. CONCLUSIONS: The proposed methodology has a similar performance to the experienced radiologists in the classification of digital mammograms into BI-RADS categories. This suggests that the pre-processing steps and modelling of each radiologist's label allows for a better estimation of the unknown ground truth labels.

Background and Objective: Breast cancer is the most frequent cancer in women. The Spanish healthcare network established population-based screening programs in all Autonomous Communities, where mammograms of asymptomatic women are taken with early diagnosis purposes. Breast density assessed from digital mammograms is a biomarker known to be related to a higher risk to develop breast cancer. It is thus crucial to provide a reliable method to measure breast density from mammograms. Furthermore the complete automation of this segmentation process is becoming fundamental as the amount of mammograms increases every day. Important challenges are related with the differences in images from different devices and the lack of an objective gold standard. This paper presents a fully automated framework based on deep learning to estimate the breast density. The framework covers breast detection, pectoral muscle exclusion, and fibroglandular tissue segmentation. Methods: A multi-center study, composed of 1785 women whose "for presentation" mammograms were segmented by two experienced radiologists. A total of 4992 of the 6680 mammograms were used as training corpus and the remaining (1688) formed the test corpus. This paper presents a histogram normalization step that smoothed the difference between acquisition, a regression architecture that learned segmentation parameters as intrinsic image features and a loss function based on the DICE score. Results: The results obtained indicate that the level of concordance (DICE score) reached by the two radiologists (0.77) was also achieved by the automated framework when it was compared to the closest breast segmentation from the radiologists. For the acquired with the highest quality device, the DICE score per acquisition device reached 0.84, while the concordance between radiologists was 0.76. Conclusions: An automatic breast density estimator based on deep learning exhibits similar performance when compared with two experienced radiologists. It suggests that this system could be used to support radiologists to ease its work. (c) 2020 Elsevier B.V. All rights reserved.

Purpose: We explored the association of the previously described Western, prudent and Mediterranean dietary patterns with prostate cancer risk by tumor aggressiveness and extension. Materials and Methods: MCC-Spain (Multicase-Control Study on Common Tumors in Spain) is a population based, multicase-control study that was done in 7 Spanish provinces between September 2008 and December 2013. It collected anthropometric, epidemiological and dietary information on 754 histologically confirmed incident cases of prostate cancer and 1,277 controls 38 to 85 years old. Three previously identified dietary patterns, including Western, prudent and Mediterranean, were reconstructed using MCC-Spain data. The association of each pattern with prostate cancer risk was assessed by logistic regression models with random, province specific intercepts. Risk according to tumor aggressiveness (Gleason score 6 vs greater than 6) and extension (cT1-cT2a vs cT2b-cT4) was evaluated by multinomial regression models. Results: High adherence to a Mediterranean dietary pattern rich not only in fruits and vegetables but also in fish, legumes and olive oil was specifically associated with a lower risk of Gleason score greater than 6 prostate cancer (quartile 3 vs 1 relative RR 0.66, 95% CI 0.46-0.96 and quartile 4 vs 1 relative RR 0.68, 95% CI 0.46-1.01, p-trend = 0.023) or with higher clinical stage (cT2b-T4 quartile 4 vs 1 relative RR 0.49, 95% CI 0.25-0.96, p-trend = 0.024). This association was not observed with the prudent pattern, which combines vegetables and fruits with low fat dairy products, whole grains and juices. The Western pattern did not show any association with prostate cancer risk. Conclusions: Nutritional recommendations for prostate cancer prevention should consider whole dietary patterns instead of individual foods. We found important differences between the Mediterranean dietary pattern, which was associated with a lower risk of aggressive prostate cancer, and Western and prudent dietary patterns, which had no relationship with prostate cancer risk.

Objectives The association between occupational exposures and mammographic density (MD), a marker of breast cancer risk, has not been previously explored. Our objective was to investigate the influence of occupational exposure to chemical, physical and microbiological agents on MD in adult women. Methods This is a population-based cross-sectional study based on 1476 female workers aged 45-65 years from seven Spanish breast cancer screening programmes. Occupational history was surveyed by trained staff. Exposure to occupational agents was assessed using the Spanish job-exposure matrix MatEmESp. Percentage of MD was measured by two radiologists using a semiautomatic computer tool. The association was estimated using mixed log-linear regression models adjusting for age, education, body mass index, menopausal status, parity, smoking, alcohol intake, type of mammography, family history of breast cancer and hormonal therapy use, and including screening centre and professional reader as random effects terms. Results Although no association was found with most of the agents, women occupationally exposed to perchloroethylene (e(beta)=1.51; 95% CI 1.04 to 2.19), ionising radiation (e(beta)=1.23; 95% CI 0.99 to 1.52) and mould spores (e(beta)=1.44; 95% CI 1.01 to 2.04) tended to have higher MD. The percentage of density increased 12% for every 5 years exposure to perchloroethylene or mould spores, 11% for every 5 years exposure to aliphatic/alicyclic hydrocarbon solvents and 3% for each 5 years exposure to ionising radiation. Conclusions Exposure to perchloroethylene, ionising radiation, mould spores or aliphatic/alicyclic hydrocarbon solvents in occupational settings could be associated with higher MD. Further studies are needed to clarify the accuracy and the reasons for these findings.

Breast cancer is the most frequent tumor in women worldwide, although well-established risk factors account for 53%-55% of cases. Therefore, other risk factors, including environmental exposures, may explain the remaining variation. Our objective was to assess the relationship between risk of breast cancer and residential proximity to industries, according to categories of industrial groups and specific pollutants released, in the context of a population-based multicase-control study of incident cancer carried out in Spain (MCC-Spain). Using the current residence of cases and controls, this study was restricted to small administrative divisions, including both breast cancer cases (452) and controls (1511) in the 10 geographical areas recruiting breast cancer cases. Distances were calculated from the respective woman's residences to the 116 industries located in the study area. We used logistic regression to estimate odds ratios (ORs) and 95% confidence intervals (95%CIs) for categories of distance (between 1 km and 3 km) to industrial plants, adjusting for matching variables and other confounders. Excess risk (OR; 95%CI) of breast cancer was found near industries overall (1.30; 1.00-1.69 at 3 km), particularly organic chemical industry (2.12; 1.20-3.76 at 2.5 km), food/beverage sector (1.87; 1.26-2.78 at 3 km), ceramic (4.71; 1.62-13.66 at 1.5 km), surface treatment with organic solvents (2.00; 1.23-3.24 at 3 km), and surface treatment of plastic and metals (1.51; 1.06-2.14 at 3 km). By pollutants, the excess risk (OR; 95% CI) was detected near industries releasing pesticides (2.09; 1.14-3.82 at 2 km), and dichloromethane (2.09; 1.28-3.40 at 3 km). Our results suggest a possible increased risk of breast cancer in women living near specific industrial plants and support the need for more detailed exposure assessment of certain agents released by these plants. (C) 2018 Elsevier Ltd. All rights reserved.

Background The influence of dietary habits on the development of gastric adenocarcinoma is not clear. The objective of the present study was to explore the association of three previously identified dietary patterns with gastric adenocarcinoma by sex, age, cancer site, and morphology. Methods MCC-Spain is a multicase-control study that included 295 incident cases of gastric adenocarcinoma and 3040 controls. The association of the Western, Prudent, and Mediterranean dietary patterns-derived in another Spanish case-control study-with gastric adenocarcinoma was assessed using multivariable logistic regression models with random province-specific intercepts and considering a possible interaction with sex and age. Risk according to tumor site (cardia, non-cardia) and morphology (intestinal/diffuse) was evaluated using multinomial regression models. Results A high adherence to the Western pattern increased gastric adenocarcinoma risk [odds ratio(fourth_vs._first_quartile) (95% confidence interval), 2.09 (1.31; 3.33)] even at low levels [odds ratio(second_vs._first_quartile) (95% confidence interval), 1.63 (1.05; 2.52)]. High adherence to the Mediterranean dietary pattern could prevent gastric adenocarcinoma [odds ratio(fourth_vs._first_quartile) (95% confidence interval), 0.53 (0.34; 0.82)]. Although no significant heterogeneity of effects was observed, the harmful effect of the Western pattern was stronger among older participants and for non-cardia adenocarcinomas, whereas the protective effect of the Mediterranean pattern was only observed among younger participants and for non-cardia tumors. Conclusion Decreasing the consumption of fatty and sugary products and of red and processed meat in favor of an increase in the intake of fruits, vegetables, legumes, olive oil, nuts, and fish might prevent gastric adenocarcinoma.

OBJECTIVE: To examine the association between two dietary patterns (Western and Mediterranean), previously linked to breast cancer risk, and mammographic density. METHODS: This cross-sectional study included 3,584 women attending population-based breast cancer screening programs and recruited between October 7, 2007, and July 14, 2008 (participation rate 74.5%). Collected data included anthropometric measurements; demographic, obstetric, and gynecologic characteristics; family and personal health history; and diet in the preceding year. Mammographic density was blindly assessed by a single radiologist and classified into four categories: less than 10%, 10-25%, 25-50%, and greater than 50%. The association between adherence to either a Western or a Mediterranean dietary pattern and mammographic density was explored using multivariable ordinal logistic regression models with random center-specific intercepts. Models were adjusted for age, body mass index, parity, menopause, smoking, family history, hormonal treatment, and calorie and alcohol intake. Differences according to women's characteristics were tested including interaction terms. RESULTS: Women with a higher adherence to the Western dietary pattern were more likely to have high mammographic density (n=242 [27%]) than women with low adherence (n=169 [19%]) with a fully adjusted odds ratio (ORQ4vsQ1) of 1.25 (95% confidence interval [CI] 1.03-1.52). This association was confined to over-weight-obese women (adjusted ORQ4vsQ1 [95% CI] 1.41 [1.13-1.76]). No association between Mediterranean dietary pattern and mammographic density was observed. CONCLUSION: The Western dietary pattern was associated with increased mammographic density among overweight-obese women. Our results might inform specific dietary recommendations for women with high mammographic density.

Background Diet is a key modifiable risk for many chronic diseases, but it remains unclear whether dietary patterns from one study sample are generalizable to other independent populations. Objective The primary objective of this study was to assess whether data-driven dietary patterns from one study sample are applicable to other populations. The secondary objective was to assess the validity of two criteria of pattern similarity. Methods Six dietary patterns-Western (n = 3), Mediterranean, Prudent, and Healthyfrom three published studies on breast cancer were reconstructed in a case-control study of 973 breast cancer patients and 973 controls. Three more internal patterns (Western, Prudent, and Mediterranean) were derived from this case-control study's own data. Statistical analysis Applicability was assessed by comparing the six reconstructed patterns with the three internal dietary patterns, using the congruence coefficient (CC) between pattern loadings. In cases where any pair met either of two commonly used criteria for declaring patterns similar (CC >= 0.85 or a statistically significant [P< 0.05] Pearson correlation), then the true similarity of those two dietary patterns was doublechecked by comparing their associations to risk for breast cancer, to assess whether those two criteria of similarity are actually reliable. Results Five of the six reconstructed dietary patterns showed high congruence (CC > 0.9) to their corresponding dietary pattern derived from the case-control study's data. Similar associations with risk for breast cancer were found in all pairs of dietary patterns that had high CC but not in all pairs of dietary patterns with statistically significant correlations. Conclusions Similar dietary patterns can be found in independent samples. The P value of a correlation coefficient is less reliable than the CC as a criterion for declaring two dietary patterns similar. This study shows that diet scores based on a particular study are generalizable to other populations.

Epidemiologic and animal data indicate that night shift work might increase the risk for breast cancer. We evaluated the association of night work with different clinical types of breast cancer in a population based case-control study (MCC-Spain study) taking into account chronotype, an individual characteristic that may relate to night shift work adaptation. Lifetime occupational history was assessed by face-to-face interviews and shift work information was available for 1708 breast cancer cases and 1778 population controls from 10 Spanish regions, enrolled from 2008 to 2013. We evaluated three shift work domains, including shift work type (permanent vs rotating), lifetime cumulative duration and frequency. We estimated odds ratios (OR) for night work compared to day work using unconditional logistic regression models adjusting for confounders. Having ever worked permanent or rotating night shift was associated with an increased risk for breast cancer compared to day workers [odds ratio (OR) 1.18; 95 % CI 0.97, 1.43]. Chronotype was differentially associated with breast cancer depending on the duration of night shift work. Risk was higher in women with invasive tumors (OR 1.23; 95 % CI 1.00, 1.51) and for estrogen and progestagen positive tumors among premenopausal women (OR 1.44; 95 % CI 1.05, 1.99). Having ever performed night shift was associated with a small increased risk for breast cancer and especially in subgroups of women with particular hormone related characteristics.

Mammographic density (MD) is an intermediate phenotype for breast cancer. Previous studies have identified genetic variants associated with MD; however, much of the genetic contribution to MD is unexplained. We conducted a two-stage genome-wide association analysis among the participants in the Determinants of Density in Mammographies in Spain study, together with a replication analysis in women from the Australian MD Twins and Sisters Study. Our discovery set covered a total of 3,351 Caucasian women aged 45 to 68 years, recruited from Spanish breast cancer screening centres. MD was blindly assessed by a single reader using Boyd's scale. A two-stage approach was employed, including a feature selection phase exploring 575,374 SNPs in 239 pairs of women with extreme phenotypes and a verification stage for the 183 selected SNPs in the remaining sample (2,873 women). Replication was conducted in 1,786 women aged 40 to 70 years old recruited via the Australian Twin Registry, where MD were measured using Cumulus-3.0, assessing 14 SNPs with a p value

Night shift work has been classified as a probable human carcinogen based on experimental studies and limited human evidence on breast cancer. Evidence on other common cancers, such as prostate cancer, is scarce. Chronotype is an individual characteristic that may relate to night work adaptation. We evaluated night shift work with relation to prostate cancer, taking into account chronotype and disease severity in a population based case-control study in Spain. We included 1,095 prostate cancer cases and 1,388 randomly selected population controls. We collected detailed information on shift schedules (permanent vs. rotating, time schedules, duration, frequency), using lifetime occupational history. Sociodemographic and lifestyle factors were assessed by face-to-face interviews and chronotype through a validated questionnaire. We used unconditional logistic regression analysis adjusting for potential confounders. Subjects who had worked at least for one year in night shift work had a slightly higher prostate cancer risk [Odds Ratio (OR) 1.14; 95%CI 0.94, 1.37] compared with never night workers; this risk increased with longer duration of exposure (28 years: OR 1.37; 95%CI 1.05, 1.81; p-trend=0.047). Risks were more pronounced for high risk tumors [D'Amico classification, Relative Risk Ratio (RRR) 1.40; 95%CI 1.05, 1.86], particularly among subjects with longer duration of exposure (28 years: RRR 1.63; 95%CI 1.08, 2.45; p-trend=0.027). Overall risk was higher among subjects with an evening chronotype, but also increased in morning chronotypes after long-term night work. In this large population based study, we found an association between night shift work and prostate cancer particularly for tumors with worse prognosis. What's new? Up to 20% of workers do night-shift work, which may increase the risk of some cancers. In this study, the authors found that long-term night-shift work was associated with an increased risk of prostate cancer and decreased survival. Overall risk was higher among workers with an evening chronotype (i.e., a preference for working in the evening vs. in the morning), but risk also increased for morning chronotypes if the duration of night-shift work increased. These results may improve our understanding of prostate cancer etiology and potential prevention strategies.