On Differentially Private 3D Medical Image Synthesis with Controllable Latent Diffusion Models

Deniz Daum, Richard Osuala, Anneliese Riess, Georgios Kaissis, Julia A. Schnabel, Maxime Di Folco

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Generally, the small size of public medical imaging datasets coupled with stringent privacy concerns, hampers the advancement of data-hungry deep learning models in medical imaging. This study addresses these challenges for 3D cardiac MRI images in the short-axis view. We propose Latent Diffusion Models that generate synthetic images conditioned on medical attributes, while ensuring patient privacy through differentially private model training. To our knowledge, this is the first work to apply and quantify differential privacy in 3D medical image generation. We pre-train our models on public data and finetune them with differential privacy on the UK Biobank dataset. Our experiments reveal that pre-training significantly improves model performance, achieving a Fréchet Inception Distance (FID) of 26.77 at ϵ=10, compared to 92.52 for models without pre-training. Additionally, we explore the trade-off between privacy constraints and image quality, investigating how tighter privacy budgets affect output controllability and may lead to degraded performance. Our results demonstrate that proper consideration during training with differential privacy can substantially improve the quality of synthetic cardiac MRI images, but there are still notable challenges in achieving consistent medical realism. Code: https://github.com/compai-lab/2024-miccai-dgm-daum.

Original languageEnglish
Title of host publicationDeep Generative Models - 4th MICCAI Workshop, DGM4MICCAI 2024, Held in Conjunction with MICCAI 2024, Proceedings
EditorsAnirban Mukhopadhyay, Ilkay Oksuz, Sandy Engelhardt, Dorit Mehrof, Yixuan Yuan
PublisherSpringer Science and Business Media Deutschland GmbH
Pages139-149
Number of pages11
ISBN (Print)9783031727436
DOIs
StatePublished - 2025
Event4th Workshop on Deep Generative Models for Medical Image Computing and Computer Assisted Intervention, DGM4MICCAI 2024, held in Conjunction with 27th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2024 - Marrakesh, Morocco
Duration: 10 Oct 202410 Oct 2024

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume15224 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference4th Workshop on Deep Generative Models for Medical Image Computing and Computer Assisted Intervention, DGM4MICCAI 2024, held in Conjunction with 27th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2024
Country/TerritoryMorocco
City Marrakesh
Period10/10/2410/10/24

Keywords

  • Cardiac MRI
  • Differential Privacy
  • Generative Models
  • Synthetic Data

Fingerprint

Dive into the research topics of 'On Differentially Private 3D Medical Image Synthesis with Controllable Latent Diffusion Models'. Together they form a unique fingerprint.

Cite this