We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

eess.IV

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Electrical Engineering and Systems Science > Image and Video Processing

Title: Towards disease-aware image editing of chest X-rays

Abstract: Disease-aware image editing by means of generative adversarial networks (GANs) constitutes a promising avenue for advancing the use of AI in the healthcare sector. Here, we present a proof of concept of this idea. While GAN-based techniques have been successful in generating and manipulating natural images, their application to the medical domain, however, is still in its infancy. Working with the CheXpert data set, we show that StyleGAN can be trained to generate realistic chest X-rays. Inspired by the Cyclic Reverse Generator (CRG) framework, we train an encoder that allows for faithfully inverting the generator on synthetic X-rays and provides organ-level reconstructions of real ones. Employing a guided manipulation of latent codes, we confer the medical condition of cardiomegaly (increased heart size) onto real X-rays from healthy patients. This work was presented in the Medical Imaging meets Neurips Workshop 2020, which was held as part of the 34th Conference on Neural Information Processing Systems (NeurIPS 2020) in Vancouver, Canada
Subjects: Image and Video Processing (eess.IV); Computer Vision and Pattern Recognition (cs.CV)
Cite as: arXiv:2109.01071 [eess.IV]
  (or arXiv:2109.01071v2 [eess.IV] for this version)

Submission history

From: Sai Niranjan Ramachandran [view email]
[v1] Thu, 2 Sep 2021 16:38:24 GMT (2323kb,D)
[v2] Fri, 3 Sep 2021 04:44:14 GMT (2324kb,D)

Link back to: arXiv, form interface, contact.