Year: 2025
Type: article
Abstract: Breast cancer has been the leading cause of cancer-related deaths among Mexican women since 2006, underscoring the need for improved diagnostic tools and accessible datasets for artificial intelligence (AI). We introduce Mammo-MX, a publicly available large-scale mammography dataset specifically focused on Mexican patients, addressing a critical gap in well-labeled data for deep learning applications. The dataset was acquired between 2023-2024 at the Jalisco Breast Clinic using a HOLOGIC ® Selenia Dimensions full-field digital mammography system, ensuring high-quality imaging. After rigorous curation, Mammo-MX comprises 13,659 mammograms from 3,368 patients with standard craniocaudal (CC) and mediolateral-oblique (MLO) views. Each study was annotated by expert radiologists according to the BI-RADS classification system, and it includes breast density assessments and extensive acquisition metadata. By offering a robust, well-labeled, and freely accessible dataset, Mammo-MX fills a void in current resources, enabling the development of AI models tailored to the Mexican population while also strengthening the global diversity and generalizability of computer-aided diagnostic tools for breast cancer.
Authors: Blanca Olivia Murillo-Ortiz, Luis Carlos Padierna, Íñigo Alonso Perea-Campos, Luis Fernando Parra-Sánchez, Sergio Meza-Chavolla, Samuel Rivera Rivera
Institutions: Mexican Social Security Institute, Universidad de Guanajuato
Language: English
Topic: AI in cancer detection
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $3,000
Award: G550401833
OpenAlex ID: W4416929952