Benign Breast Surgical Biopsies: Are They Always Justified?

Active Learning in Context-Driven Stream Mining With an Application to Image Mining

We propose an image stream mining method in which images arrive with contexts (metadata) and need to be processed in real time by the image mining system (IMS), which needs to make predictions and derive actionable intelligence from these streams. After extracting the features of the image by preprocessing, IMS determines online the classifier to use on the extracted features to make a prediction using the context of the image. A key challenge associated with stream mining is that the prediction accuracy of the classifiers is unknown, since the image source is unknown; thus, these accuracies need to be learned online. Another key challenge of stream mining is that learning can only be done by observing the true label, but this is costly to obtain. To address these challenges, we model the image stream mining problem as an active, online contextual experts problem, where the context of the image is used to guide the classifier selection decision. We develop an active learning algorithm and show that it achieves regret sublinear in the number of images that have been observed so far. To further illustrate and assess the performance of our proposed methods, we apply them to diagnose breast cancer from the images of cellular samples obtained from the fine needle aspirate of breast mass. Our findings show that very high diagnosis accuracy can be achieved by actively obtaining only a small fraction of true labels through surgical biopsies. Other applications include video surveillance and video traffic monitoring.

Interdependence of radial scar and proliferative disease with respect to invasive breast carcinoma risk in patients with benign breast biopsies

BACKGROUND

Radial scars (RS) are benign breast lesions that have been implicated as independent risk factors for invasive breast carcinoma (IBC).

METHODS

A retrospective cohort study of 9556 women who underwent biopsy between 1950-1986 and enrolled in the Nashville Breast Cohort was performed to investigate the association between RS in a benign breast biopsy and the risk of IBC. The risk associated with RS and coexistent proliferative disease (PD) was assessed adjusting for age at biopsy using a Cox hazards regression analysis with time-dependent covariates.

RESULTS

RS were identified in 880 women (9.2%). The average follow-up time was 20.4 years. Sixty-two women (7.0%) with RS developed IBC compared with 5.5% of controls. The relative risk of IBC associated with RS was 1.82 (95% confidence interval [95% CI], 1.2-2.7) at 10 years. Restricting the analysis to women age > 49 years increased the risk to 2.14 (95% CI, 0.6-2.8). These risks decreased with increasing years of follow-up. Approximately 92% of women with RS also had PD, but RS were present in only 1.3% of biopsies without PD. Analyses stratifying relative risk with regard to PD found RS to minimally elevate the relative risk of subsequent IBC.

CONCLUSIONS

RS in the absence of PD is uncommon. Although the presence of RS in a benign breast biopsy mildly elevates the risk of IBC risk, the current analysis indicated that this risk can be largely attributed to the category of coexistent PD. In women with both RS and atypical hyperplasia, recommendations for interventions beyond biopsy should be based on the extent of atypical hyperplasia.