Artificial Intelligence–Assisted Detection of Breast Cancer Lymph Node Metastases in the Post-Neoadjuvant Treatment Setting
Bibliographic record
Abstract
Lymph node assessment for metastasis is a common, time-consuming, and potentially error-prone pathologist task. Past studies have proposed deep learning algorithms designed to automate this task. However, none have explicitly evaluated the generalizability of these algorithms to lymph node in patients with breast cancer who have received neoadjuvant systemic therapy (NAT). In this study, we created a large 1027-slide data set exclusively containing patients with breast cancer who have received NAT with detailed pathologist labels. We developed an interpretable deep learning pipeline to carry out the following 2 tasks: first, to classify slides as positive or negative for metastases, and second, to create a detailed, patch-level heatmap for probability of metastasis. We evaluated this pipeline with and without post-NAT treatment effect in training data, and investigated its performance relative to both slide- and patch-level tasks. We found 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 for slide-level classification, and 70.9% specificity when calibrating for 100% sensitivity. We additionally performed an interpretability study on the outputs of our pipeline and found that the patch-level heatmap was successful in efficiently guiding pathologists toward detecting and correcting erroneous predictions that were made with an uncalibrated network.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.002 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".