Cell Anomaly Localisation using Structured Uncertainty Prediction Networks

Boyko Vodenicharski, Samuel McDermott, Katherine M. Webber, Viola Introini, Pietro Cicuta, Richard Bowman, Ivor J. A. Simpson, Neill D. F. Campbell
Proceedings of The 5th International Conference on Medical Imaging with Deep Learning, PMLR 172:1285-1300, 2022.

Abstract

This paper proposes an unsupervised approach to anomaly detection in bright-field or fluorescence cell microscopy, where our goal is to localise malaria parasites. This is achieved by building a generative model (a variational autoencoder) that describes healthy cell images, where we additionally model the structure of the predicted image uncertainty, rather than assuming pixelwise independence in the likelihood function. This provides a whitened residual representation, where the anticipated structured mistakes by the generative model are reduced, but distinctive structures that did not occur in the training distribution, e.g. parasites are highlighted. We employ the recently published Structured Uncertainty Prediction Networks approach to enable tractable learning of the uncertainty structure. Here, the residual covariance matrix is efficiently approximated using a sparse Cholesky parameterisation. We demonstrate that our proposed approach is more effective for detecting real and synthetic structured image perturbations compared to diagonal Gaussian likelihoods.

Cite this Paper

BibTeX
@InProceedings{pmlr-v172-vodenicharski22a, title = {Cell Anomaly Localisation using Structured Uncertainty Prediction Networks}, author = {Vodenicharski, Boyko and McDermott, Samuel and Webber, Katherine M. and Introini, Viola and Cicuta, Pietro and Bowman, Richard and Simpson, Ivor J. A. and Campbell, Neill D. F.}, booktitle = {Proceedings of The 5th International Conference on Medical Imaging with Deep Learning}, pages = {1285--1300}, 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/vodenicharski22a/vodenicharski22a.pdf}, url = {https://proceedings.mlr.press/v172/vodenicharski22a.html}, abstract = {This paper proposes an unsupervised approach to anomaly detection in bright-field or fluorescence cell microscopy, where our goal is to localise malaria parasites. This is achieved by building a generative model (a variational autoencoder) that describes healthy cell images, where we additionally model the structure of the predicted image uncertainty, rather than assuming pixelwise independence in the likelihood function. This provides a whitened residual representation, where the anticipated structured mistakes by the generative model are reduced, but distinctive structures that did not occur in the training distribution, e.g. parasites are highlighted. We employ the recently published Structured Uncertainty Prediction Networks approach to enable tractable learning of the uncertainty structure. Here, the residual covariance matrix is efficiently approximated using a sparse Cholesky parameterisation. We demonstrate that our proposed approach is more effective for detecting real and synthetic structured image perturbations compared to diagonal Gaussian likelihoods.} }
Endnote
%0 Conference Paper %T Cell Anomaly Localisation using Structured Uncertainty Prediction Networks %A Boyko Vodenicharski %A Samuel McDermott %A Katherine M. Webber %A Viola Introini %A Pietro Cicuta %A Richard Bowman %A Ivor J. A. Simpson %A Neill D. F. Campbell %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-vodenicharski22a %I PMLR %P 1285--1300 %U https://proceedings.mlr.press/v172/vodenicharski22a.html %V 172 %X This paper proposes an unsupervised approach to anomaly detection in bright-field or fluorescence cell microscopy, where our goal is to localise malaria parasites. This is achieved by building a generative model (a variational autoencoder) that describes healthy cell images, where we additionally model the structure of the predicted image uncertainty, rather than assuming pixelwise independence in the likelihood function. This provides a whitened residual representation, where the anticipated structured mistakes by the generative model are reduced, but distinctive structures that did not occur in the training distribution, e.g. parasites are highlighted. We employ the recently published Structured Uncertainty Prediction Networks approach to enable tractable learning of the uncertainty structure. Here, the residual covariance matrix is efficiently approximated using a sparse Cholesky parameterisation. We demonstrate that our proposed approach is more effective for detecting real and synthetic structured image perturbations compared to diagonal Gaussian likelihoods.
APA
Vodenicharski, B., McDermott, S., Webber, K.M., Introini, V., Cicuta, P., Bowman, R., Simpson, I.J.A. & Campbell, N.D.F.. (2022). Cell Anomaly Localisation using Structured Uncertainty Prediction Networks. Proceedings of The 5th International Conference on Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 172:1285-1300 Available from https://proceedings.mlr.press/v172/vodenicharski22a.html.

Related Material