AI-assisted detection of breast cancer lymph node metastases in the post-neoadjuvant treatment setting.

Abstract

Lymph node assessment for metastasis is a common, time-consuming, and potentially error prone pathologist task. Past studies have proposed deep learning (DL) algorithms designed to automate this task. However, none have explicitly evaluated the generalizability of these algorithms to lymph nodes in breast cancer patients who have received post-neoadjuvant systemic therapy (NAT). In this study, we create a large, 1027-slide dataset containing exclusively post-NAT breast cancer patients with detailed pathologist labels. We develop an interpretable DL pipeline to carry out two tasks: firstly, to classify slides as positive or negative for metastases, and secondly, to create a detailed, patch-level heatmap for probability of metastasis. We evaluate this pipeline with and without post-NAT treatment effect in training data, and investigate its performance relative to both slide- and patch-level tasks. We find that the presence of post-NAT treatment effect training data is relevant for both tasks, with particular benefits in pipeline specificity. With the post-NAT testing cohort, we found that our final pipeline obtained 0.986 area under the receiver operating characteristic curve (AUC) for slide-level classification, and 70.9% specificity when calibrating for 100% sensitivity. We additionally perform an interpretability study on the outputs of our pipeline, and find that the patch-level heatmap was successful in efficiently guiding pathologists towards detecting and correcting erroneous predictions that were made with an uncalibrated network.