Interpreting Latent Spaces of Generative Models for Medical Images using Unsupervised Methods

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review


  • Fulltext

    Submitted manuscript, 21.5 MB, PDF document

Generative models such as Generative Adversarial Networks (GANs) and Variational Autoencoders (VAEs) play an increasingly important role in medical image analysis. The latent spaces of these models often show semantically meaningful directions corresponding to human-interpretable image transformations. However, until now, their exploration for medical images has been limited due to the requirement of supervised data. Several methods for unsupervised discovery of interpretable directions in GAN latent spaces have shown interesting results on natural images. This work explores the potential of applying these techniques on medical images by training a GAN and a VAE on thoracic CT scans and using an unsupervised method to discover interpretable directions in the resulting latent space. We find several directions corresponding to non-trivial image transformations, such as rotation or breast size. Furthermore, the directions show that the generative models capture 3D structure despite being presented only with 2D data. The results show that unsupervised methods to discover interpretable directions in GANs generalize to VAEs and can be applied to medical images. This opens a wide array of future work using these methods in medical image analysis.
Original languageEnglish
Title of host publicationDeep Generative Models : Second MICCAI Workshop, DGM4MICCAI 2022, Held in Conjunction with MICCAI 2022, Singapore, September 22, 2022, Proceedings
Publication date2022
ISBN (Print)978-3-031-18575-5
ISBN (Electronic)978-3-031-18576-2
Publication statusPublished - 2022
EventSecond MICCAI Workshop, DGM4MICCAI 2022, - Singapore
Duration: 22 Sep 2022 → …


WorkshopSecond MICCAI Workshop, DGM4MICCAI 2022,
Periode22/09/2022 → …
SeriesLecture Notes in Computer Science

Number of downloads are based on statistics from Google Scholar and

No data available

ID: 335278278