Image-conditioned diffusion models for medical anomaly detection
File(s)
Accepted version
Author(s)
Type
Chapter
Abstract
Generating pseudo-healthy reconstructions of images is an effective way to detect anomalies, as identifying the differences between the reconstruction and the original can localise arbitrary anomalies whilst also providing interpretability for an observer by displaying what the image ‘should’ look like. All existing reconstruction-based methods have a common shortcoming; they assume that models trained on purely normal data are incapable of reproducing pathologies yet also able to fully maintain healthy tissue. These implicit assumptions often fail, with models either not recovering normal regions or reproducing both the normal and abnormal features. We rectify this issue using image-conditioned diffusion models. Our model takes the input image as conditioning and is explicitly trained to correct synthetic anomalies introduced into healthy images, ensuring that it removes anomalies at test time. This conditioning allows the model to attend to the entire image without any loss of information, enabling it to replicate healthy regions with high fidelity. We evaluate our method across four datasets and define a new state-of-the-art performance for residual-based anomaly detection. Code is available at https://github.com/matt-baugh/img-cond-diffusion-model-ad.
Editor(s)
Sudre, CH
Mehta, R
Ouyang, C
Qin, C
Rakic, M
Wells, WM
Date Issued
2024-10-03
Citation
Uncertainty for Safe Utilization of Machine Learning in Medical Imaging, 2024, 15167, pp.117-127
ISBN
978-3-031-73157-0
Publisher
Springer Nature Switzerland AG
Start Page
117
End Page
127
Journal / Book Title
Uncertainty for Safe Utilization of Machine Learning in Medical Imaging
Lecture Notes in Computer Science
Volume
15167
Copyright Statement
© 2025 The Author(s), under exclusive license to Springer Nature Switzerland AG. This is the author’s accepted manuscript made available under a CC-BY licence in accordance with Imperial’s Research Publications Open Access policy (www.imperial.ac.uk/oa-policy)
License URL
Subjects
Anomaly detection
Computer Science
Computer Science, Artificial Intelligence
Computer Science, Theory & Methods
Diffusion model
Life Sciences & Biomedicine
Radiology, Nuclear Medicine & Medical Imaging
Science & Technology
Self-supervised
Technology
Publication Status
Published
Date Publish Online
2024-10-03