Denoising Autoencoders for Unsupervised Anomaly Detection in Brain MRI

Antanas Kascenas, Nicolas Pugeault, Alison Q. O’Neil
Proceedings of The 5th International Conference on Medical Imaging with Deep Learning, PMLR 172:653-664, 2022.

Abstract

Pathological brain lesions exhibit diverse appearance in brain images, making it difficult to train supervised detection solutions due to the lack of comprehensive data and annotations. Thus, in this work we tackle unsupervised anomaly detection, using only healthy data for training with the aim of detecting unseen anomalies at test time. Many current approaches employ autoencoders with restrictive architectures (i.e. containing information bottlenecks) that tend to give poor reconstructions of not only the anomalous but also the normal parts of the brain. Instead, we investigate classical denoising autoencoder models that do not require bottlenecks and can employ skip connections to give high resolution fidelity. We design a simple noise generation method of upscaling low-resolution noise that enables high-quality reconstructions. We find that with appropriate noise generation, denoising autoencoder reconstruction errors generalize to hyperintense lesion segmentation and reach state of the art performance for unsupervised tumor detection in brain MRI data, beating more complex methods such as variational autoencoders. We believe this provides a strong and easy-to-implement baseline for further research into unsupervised anomaly detection.

Cite this Paper

BibTeX
@InProceedings{pmlr-v172-kascenas22a, title = {Denoising Autoencoders for Unsupervised Anomaly Detection in Brain MRI}, author = {Kascenas, Antanas and Pugeault, Nicolas and O'Neil, Alison Q.}, booktitle = {Proceedings of The 5th International Conference on Medical Imaging with Deep Learning}, pages = {653--664}, year = {2022}, editor = {Konukoglu, Ender and Menze, Bjoern and Venkataraman, Archana and Baumgartner, Christian and Dou, Qi and Albarqouni, Shadi}, volume = {172}, series = {Proceedings of Machine Learning Research}, month = {06--08 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v172/kascenas22a/kascenas22a.pdf}, url = {https://proceedings.mlr.press/v172/kascenas22a.html}, abstract = {Pathological brain lesions exhibit diverse appearance in brain images, making it difficult to train supervised detection solutions due to the lack of comprehensive data and annotations. Thus, in this work we tackle unsupervised anomaly detection, using only healthy data for training with the aim of detecting unseen anomalies at test time. Many current approaches employ autoencoders with restrictive architectures (i.e. containing information bottlenecks) that tend to give poor reconstructions of not only the anomalous but also the normal parts of the brain. Instead, we investigate classical denoising autoencoder models that do not require bottlenecks and can employ skip connections to give high resolution fidelity. We design a simple noise generation method of upscaling low-resolution noise that enables high-quality reconstructions. We find that with appropriate noise generation, denoising autoencoder reconstruction errors generalize to hyperintense lesion segmentation and reach state of the art performance for unsupervised tumor detection in brain MRI data, beating more complex methods such as variational autoencoders. We believe this provides a strong and easy-to-implement baseline for further research into unsupervised anomaly detection.} }
Endnote
%0 Conference Paper %T Denoising Autoencoders for Unsupervised Anomaly Detection in Brain MRI %A Antanas Kascenas %A Nicolas Pugeault %A Alison Q. O’Neil %B Proceedings of The 5th International Conference on Medical Imaging with Deep Learning %C Proceedings of Machine Learning Research %D 2022 %E Ender Konukoglu %E Bjoern Menze %E Archana Venkataraman %E Christian Baumgartner %E Qi Dou %E Shadi Albarqouni %F pmlr-v172-kascenas22a %I PMLR %P 653--664 %U https://proceedings.mlr.press/v172/kascenas22a.html %V 172 %X Pathological brain lesions exhibit diverse appearance in brain images, making it difficult to train supervised detection solutions due to the lack of comprehensive data and annotations. Thus, in this work we tackle unsupervised anomaly detection, using only healthy data for training with the aim of detecting unseen anomalies at test time. Many current approaches employ autoencoders with restrictive architectures (i.e. containing information bottlenecks) that tend to give poor reconstructions of not only the anomalous but also the normal parts of the brain. Instead, we investigate classical denoising autoencoder models that do not require bottlenecks and can employ skip connections to give high resolution fidelity. We design a simple noise generation method of upscaling low-resolution noise that enables high-quality reconstructions. We find that with appropriate noise generation, denoising autoencoder reconstruction errors generalize to hyperintense lesion segmentation and reach state of the art performance for unsupervised tumor detection in brain MRI data, beating more complex methods such as variational autoencoders. We believe this provides a strong and easy-to-implement baseline for further research into unsupervised anomaly detection.
APA
Kascenas, A., Pugeault, N. & O’Neil, A.Q.. (2022). Denoising Autoencoders for Unsupervised Anomaly Detection in Brain MRI. Proceedings of The 5th International Conference on Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 172:653-664 Available from https://proceedings.mlr.press/v172/kascenas22a.html.

Related Material