Skip to main navigation Skip to search Skip to main content

DeepMammo: Deep Learning algorithm for Digital Mammogram Source Identification

Research output: Contribution to journalConference articlepeer-review

Abstract

Advances in AI allow for fake image creation. These techniques can be used to fake mammograms. This could impact patient care and medicolegal cases. One method to verify that an image is original is to confirm the source of the image. A deep-learning algorithm(DeepMammo)-based on CNNs and FCNNs, used to identify the machine that created any mammogram. We analyze mammograms of 1574 patients obtained on 7-different mammography machines and randomly split the dataset by patient into training/validation(80%) and test(20%) datasets. DeepMammo has an accuracy of 98.09%, AUC of 95.96% in the test dataset.

Original languageEnglish (US)
Article number342
JournalIS and T International Symposium on Electronic Imaging Science and Technology
Volume36
Issue number4
DOIs
StatePublished - 2024
Externally publishedYes
EventIS and T International Symposium on Electronic Imaging 2024: Media Watermarking, Security, and Forensics, MWSF 2024 - San Francisco, United States
Duration: Jan 21 2024Jan 25 2024

Keywords

  • Deep Learning
  • Forensic Science
  • Machine Learning
  • Mammograms
  • Medical Imaging
  • Radiology

ASJC Scopus subject areas

  • Computer Graphics and Computer-Aided Design
  • Computer Science Applications
  • Human-Computer Interaction
  • Software
  • Electrical and Electronic Engineering
  • Atomic and Molecular Physics, and Optics

Fingerprint

Dive into the research topics of 'DeepMammo: Deep Learning algorithm for Digital Mammogram Source Identification'. Together they form a unique fingerprint.

Cite this