Impact of multi-source data augmentation on performance of convolutional neural networks for abnormality classification in mammography

In Chan Hwang, Hari Trivedi, Beatrice Brown-Mulry, Linglin Zhang, Vineela Nalla, Aimilia Gastounioti, Judy Gichoya, Laleh Seyyed-Kalantari, Imon Banerjee, Min Jae Woo

Research output: Contribution to journalArticlepeer-review

Abstract

Introduction: To date, most mammography-related AI models have been trained using either film or digital mammogram datasets with little overlap. We investigated whether or not combining film and digital mammography during training will help or hinder modern models designed for use on digital mammograms. Methods: To this end, a total of six binary classifiers were trained for comparison. The first three classifiers were trained using images only from Emory Breast Imaging Dataset (EMBED) using ResNet50, ResNet101, and ResNet152 architectures. The next three classifiers were trained using images from EMBED, Curated Breast Imaging Subset of Digital Database for Screening Mammography (CBIS-DDSM), and Digital Database for Screening Mammography (DDSM) datasets. All six models were tested only on digital mammograms from EMBED. Results: The results showed that performance degradation to the customized ResNet models was statistically significant overall when EMBED dataset was augmented with CBIS-DDSM/DDSM. While the performance degradation was observed in all racial subgroups, some races are subject to more severe performance drop as compared to other races. Discussion: The degradation may potentially be due to (1) a mismatch in features between film-based and digital mammograms (2) a mismatch in pathologic and radiological information. In conclusion, use of both film and digital mammography during training may hinder modern models designed for breast cancer screening. Caution is required when combining film-based and digital mammograms or when utilizing pathologic and radiological information simultaneously.

Original languageEnglish
Article number1181190
JournalFrontiers in Radiology
Volume3
DOIs
StatePublished - 2023

Keywords

  • breast cancer
  • cancer screening (MeSH)
  • CBIS-DDSM
  • deep learning—artificial intelligence
  • EMBED
  • FFDM—full field digital mammography
  • mammography

Fingerprint

Dive into the research topics of 'Impact of multi-source data augmentation on performance of convolutional neural networks for abnormality classification in mammography'. Together they form a unique fingerprint.

Cite this