Faculty Bibliography

A 50-year-old woman with persistent axillary lymphadenopathy 17 weeks following COVID-19 vaccination was ultimately diagnosed with biopsy-proven benign reactive lymphadenopathy. In contrast, a 60-year-old woman with axillary lymphadenopathy and concurrent suspicious breast findings 9 weeks following COVID-19 vaccination was ultimately diagnosed with biopsy-proven metastatic breast carcinoma. This article reviews the current guidelines regarding breast cancer screening and management of axillary lymphadenopathy in the setting of COVID-19 vaccination.

Though consistently shown to detect mammographically occult cancers, breast ultrasound has been noted to have high false-positive rates. In this work, we present an AI system that achieves radiologist-level accuracy in identifying breast cancer in ultrasound images. Developed on 288,767 exams, consisting of 5,442,907 B-mode and Color Doppler images, the AI achieves an area under the receiver operating characteristic curve (AUROC) of 0.976 on a test set consisting of 44,755 exams. In a retrospective reader study, the AI achieves a higher AUROC than the average of ten board-certified breast radiologists (AUROC: 0.962 AI, 0.924 ± 0.02 radiologists). With the help of the AI, radiologists decrease their false positive rates by 37.3% and reduce requested biopsies by 27.8%, while maintaining the same level of sensitivity. This highlights the potential of AI in improving the accuracy, consistency, and efficiency of breast ultrasound diagnosis.

We present a deep convolutional neural network for breast cancer screening exam classification, trained and evaluated on over 200,000 exams (over 1,000,000 images). Our network achieves an AUC of 0.895 in predicting the presence of cancer in the breast, when tested on the screening population. We attribute the high accuracy to a few technical advances. (i) Our network's novel two-stage architecture and training procedure, which allows us to use a high-capacity patch-level network to learn from pixel-level labels alongside a network learning from macroscopic breast-level labels. (ii) A custom ResNet-based network used as a building block of our model, whose balance of depth and width is optimized for high-resolution medical images. (iii) Pretraining the network on screening BI-RADS classification, a related task with more noisy labels. (iv) Combining multiple input views in an optimal way among a number of possible choices. To validate our model, we conducted a reader study with 14 readers, each reading 720 screening mammogram exams, and show that our model is as accurate as experienced radiologists when presented with the same data. We also show that a hybrid model, averaging the probability of malignancy predicted by a radiologist with a prediction of our neural network, is more accurate than either of the two separately. To further understand our results, we conduct a thorough analysis of our network's performance on different subpopulations of the screening population, the model's design, training procedure, errors, and properties of its internal representations. Our best models are publicly available at https://github.com/nyukat/breastcancerclassifier.

OBJECTIVE: Antibodies to nuclear antigens (ANA) are diagnostic in several autoimmune disorders, yet failure to achieve B cell tolerance in these diseases is still poorly understood. Although secreted ANA detected by an indirect immunofluorescent assay (IFA) are the gold standard for autoreactivity, there has been no convenient assay to measure the frequency of circulating B cells that recognize nuclear antigens (ANA+ B cells) in patients. Our aim was to generate an assay to easily identify these B cells and to examine its utility in a study of autoreactive B cells in Systemic Lupus Erythematosus (SLE). METHODS: We developed and validated a novel flow cytometry-based assay that identifies ANA+ B cells using biotinylated nuclear extracts, and used it to examine B cell tolerance checkpoints in peripheral blood mononuclear cells (PBMCs) obtained from SLE patients and healthy controls (HCs). RESULTS: We show progressive selection against ANA+ B cells as they mature from transitional to naive to CD27+ IgD- and CD27+ IgD+ memory cells in both healthy subjects and SLE patients; however SLE patients failed to anergize ANA+ naive B cells to the same extent as healthy individuals. We also show anergy induction is restored in SLE patients treated with belimumab, an inhibitor of B cell-activating factor (BAFF). CONCLUSION: This assay will enable studies of large populations to identify potential genetic or environmental factors affecting B cell tolerance checkpoints in health and disease and permit monitoring of the B cell response to therapeutic interventions