Deep learning in medical imaging: Risks to patient privacy and possible solutions
University of Calgary
Nils D. Forkert, Anup Tuladhar and Matthias Wilms
Recent research in computer vision has shown that original images used for training of deep learning models can be reconstructed using so-called inversion attacks. However, the feasibility of this attack type has not been investigated for complex 3D medical images. Thus, the aim of this study was to examine the vulnerability of deep learning techniques used in medical imaging to model inversion attacks and investigate multiple quantitative metrics to evaluate the quality of the reconstructed images.
For the development and evaluation of model inversion attacks, the public LPBA40 database consisting of 40 brain magnetic resonance imaging scans with corresponding segmentations of the gyri and deep grey matter brain structures were used to train two popular deep convolutional neural networks, namely a U-Net and SegNet, and corresponding inversion decoders. Matthews correlation coefficient, the structural similarity index measure (SSIM), and the magnitude of the deformation field resulting from non-linear registration of the original and reconstructed images were used to evaluate the reconstruction accuracy.
A comparison of the similarity metrics revealed that the SSIM is best suited to evaluate the reconstruction accuracy, followed closely by the magnitude of the deformation field. The qualitative evaluation showed that training images can be reconstructed with some degradation due to blurring but can be correctly matched to the original images in the majority of the cases. In conclusion, the results of this study indicate that it is possible to reconstruct patient data used for training of convolutional neural networks and that the SSIM is a good metric to assess the reconstruction accuracy.
Project deliverables are available in the following language(s)
OPC Funded Project
This project received funding support through the Office of the Privacy Commissioner of Canada’s Contributions Program. The opinions expressed in the summary and report(s) are those of the authors and do not necessarily reflect those of the Office of the Privacy Commissioner of Canada. Summaries have been provided by the project authors. Please note that the projects appear in their language of origin.
Dr. rer nat Nils Daniel Forkert
Canada Research Chair in Medical Image Analysis
Program Director - Child Health Data Science, Alberta Children’s Hospital Research Institute
Theme Lead Machine Learning in Neuroscience, Hotchkiss Brain Institute
Associate Professor in the Departments of Radiology & Clinical Neurosciences,
Cumming School of Medicine, and Department of Electrical and Software Engineering,
Schulich School of Engineering, University of Calgary
Report a problem or mistake on this page
- Date modified: