Cross-Domain Transfer Learning Architecture for Microcalcification Cluster Detection Using the MEXBreast Multiresolution Mammography Dataset

Microcalcification clusters (MCCs) are key indicators of breast cancer, with studies showing that approximately 50% of mammograms with MCCs confirm a cancer diagnosis. Early detection is critical, as it ensures a five-year survival rate of up to 99%. However, MCC detection remains challenging due to...

Full description

Saved in:
Bibliographic Details
Main Authors: Ricardo Salvador Luna Lozoya, Humberto de Jesús Ochoa Domínguez, Juan Humberto Sossa Azuela, Vianey Guadalupe Cruz Sánchez, Osslan Osiris Vergara Villegas, Karina Núñez Barragán
Format: Article
Language:English
Published: MDPI AG 2025-07-01
Series:Mathematics
Subjects:
Online Access:https://www.mdpi.com/2227-7390/13/15/2422
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Microcalcification clusters (MCCs) are key indicators of breast cancer, with studies showing that approximately 50% of mammograms with MCCs confirm a cancer diagnosis. Early detection is critical, as it ensures a five-year survival rate of up to 99%. However, MCC detection remains challenging due to their features, such as small size, texture, shape, and impalpability. Convolutional neural networks (CNNs) offer a solution for MCC detection. Nevertheless, CNNs are typically trained on single-resolution images, limiting their generalizability across different image resolutions. We propose a CNN trained on digital mammograms with three common resolutions: 50, 70, and 100 <inline-formula><math display="inline"><semantics><mi mathvariant="sans-serif">μ</mi></semantics></math></inline-formula>m. The architecture processes individual 1 cm<sup>2</sup> patches extracted from the mammograms as input samples and includes a MobileNetV2 backbone, followed by a flattening layer, a dense layer, and a sigmoid activation function. This architecture was trained to detect MCCs using patches extracted from the INbreast database, which has a resolution of 70 <inline-formula><math display="inline"><semantics><mi mathvariant="sans-serif">μ</mi></semantics></math></inline-formula>m, and achieved an accuracy of 99.84%. We applied transfer learning (TL) and trained on 50, 70, and 100 <inline-formula><math display="inline"><semantics><mi mathvariant="sans-serif">μ</mi></semantics></math></inline-formula>m resolution patches from the MEXBreast database, achieving accuracies of 98.32%, 99.27%, and 89.17%, respectively. For comparison purposes, models trained from scratch, without leveraging knowledge from the pretrained model, achieved 96.07%, 99.20%, and 83.59% accuracy for 50, 70, and 100 <inline-formula><math display="inline"><semantics><mi mathvariant="sans-serif">μ</mi></semantics></math></inline-formula>m, respectively. Results demonstrate that TL improves MCC detection across resolutions by reusing pretrained knowledge.
ISSN:2227-7390