A Multi-Institutional Open-Source Benchmark Dataset for Breast Cancer Clinical Decision Support using Synthetic Correlated Diffusion Imaging Data

TL;DR

This paper introduces Cancer-Net BCa, a comprehensive open-source dataset of synthetic correlated diffusion MRI images for breast cancer, enabling improved machine learning-based clinical decision support.

Contribution

It provides the first large, multi-institutional CDI$^s$ dataset for breast cancer, including detailed annotations and diversity analysis, to facilitate AI research in this domain.

Findings

Dataset includes 253 patients from 10 institutions.

Contains detailed annotations like lesion type, genetic subtype, and treatment response.

Analyzes demographic and tumor diversity to identify potential biases.

Abstract

Recently, a new form of magnetic resonance imaging (MRI) called synthetic correlated diffusion (CDI$^s$) imaging was introduced and showed considerable promise for clinical decision support for cancers such as prostate cancer when compared to current gold-standard MRI techniques. However, the efficacy for CDI$^s$ for other forms of cancers such as breast cancer has not been as well-explored nor have CDI$^s$ data been previously made publicly available. Motivated to advance efforts in the development of computer-aided clinical decision support for breast cancer using CDI$^s$, we introduce Cancer-Net BCa, a multi-institutional open-source benchmark dataset of volumetric CDI$^s$ imaging data of breast cancer patients. Cancer-Net BCa contains CDI$^s$ volumetric images from a pre-treatment cohort of 253 patients across ten institutions, along with detailed annotation metadata (the lesion type, genetic subtype, longest diameter on the MRI (MRLD), the Scarff-Bloom-Richardson (SBR) grade, and the post-treatment breast cancer pathologic complete response (pCR) to neoadjuvant chemotherapy). We further examine the demographic and tumour diversity of the Cancer-Net BCa dataset to gain deeper insights into potential biases. Cancer-Net BCa is publicly available as a part of a global open-source initiative dedicated to accelerating advancement in machine learning to aid clinicians in the fight against cancer.