Learning from Lost Provenance: Multiple Instance Learning for Cancer Registry Tumor Group Classification
2026-07-03 • Computation and Language
Computation and Language
AI summaryⓘ
The authors found a way to train deep learning models to classify cancer tumor groups using labels already available from cancer registries, but without needing detailed labels for each individual pathology report. They used a technique called Attention-Based Multiple Instance Learning (ABMIL) to figure out which reports mattered most for each patient-level label. This method helped create a cleaner, more useful training dataset from noisy data. Their final model performed better than previous methods and can help automate cancer registry tasks without extra manual labeling or heavy computing resources.
deep learningcancer registriespathology reportsAttention-Based Multiple Instance Learningtumor classificationpatient-level labelsmachine learningtraining dataF1 score
Authors
Leonard Ruocco, Jonathan Simkin, Lovedeep Gondora, Gregory Arbour, Raymond Ng
Abstract
Modernizing cancer registries with deep learning is opening new opportunities to automate labor-intensive tasks such as the coding of pathology reports. However, progress is constrained by the scarcity of report-level human-annotated training data. Cancer registries generate substantial volumes of expert-assigned labels as a routine product of their operations, but these exist at the patient level and are not linked to the individual pathology reports that informed them, limiting their direct use for training models. We develop an efficient framework for training deep learning classifiers by leveraging these operationally-generated labels without requiring per-report human annotation, demonstrated for tumor group classification at the BC Cancer Registry. We use Attention-Based Multiple Instance Learning (ABMIL) to recover the lost link between patient-level labels and the reports that informed them, leveraging the attention the model places on each report to distil a large, noisily-labeled corpus into a compact, high-quality per-report training dataset. A classifier fine-tuned on a distilled dataset achieved a macro F1 of 0.83, outperforming established baselines across most tumor groups. By turning routine operational labels into high-quality training data without additional annotation or large-scale computing infrastructure, ABMIL offers a practical and accessible route to automating cancer registry workflows.