Hierarchical Conditional Random Fields for Outlier Detection: An Application to Detecting Epileptogenic Cortical Malformations

Bilal Ahmed, Thomas Thesen, Karen Blackmon, Yijun Zhao, Orrin Devinsky, Ruben Kuzniecky, Carla Brodley
Proceedings of the 31st International Conference on Machine Learning, PMLR 32(2):1080-1088, 2014.

Abstract

We cast the problem of detecting and isolating regions of abnormal cortical tissue in the MRIs of epilepsy patients in an image segmentation framework. Employing a multiscale approach we divide the surface images into segments of different sizes and then classify each segment as being an outlier, by comparing it to the same region across controls. The final classification is obtained by fusing the outlier probabilities obtained at multiple scales using a tree-structured hierarchical conditional random field (HCRF). The proposed method correctly detects abnormal regions in 90% of patients whose abnormality was detected via routine visual inspection of their clinical MRI. More importantly, it detects abnormalities in 80% of patients whose abnormality escaped visual inspection by expert radiologists.

Cite this Paper

BibTeX
@InProceedings{pmlr-v32-ahmed14, title = {Hierarchical Conditional Random Fields for Outlier Detection: An Application to Detecting Epileptogenic Cortical Malformations}, author = {Ahmed, Bilal and Thesen, Thomas and Blackmon, Karen and Zhao, Yijun and Devinsky, Orrin and Kuzniecky, Ruben and Brodley, Carla}, booktitle = {Proceedings of the 31st International Conference on Machine Learning}, pages = {1080--1088}, year = {2014}, editor = {Xing, Eric P. and Jebara, Tony}, volume = {32}, number = {2}, series = {Proceedings of Machine Learning Research}, address = {Bejing, China}, month = {22--24 Jun}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v32/ahmed14.pdf}, url = {https://proceedings.mlr.press/v32/ahmed14.html}, abstract = {We cast the problem of detecting and isolating regions of abnormal cortical tissue in the MRIs of epilepsy patients in an image segmentation framework. Employing a multiscale approach we divide the surface images into segments of different sizes and then classify each segment as being an outlier, by comparing it to the same region across controls. The final classification is obtained by fusing the outlier probabilities obtained at multiple scales using a tree-structured hierarchical conditional random field (HCRF). The proposed method correctly detects abnormal regions in 90% of patients whose abnormality was detected via routine visual inspection of their clinical MRI. More importantly, it detects abnormalities in 80% of patients whose abnormality escaped visual inspection by expert radiologists.} }
Endnote
%0 Conference Paper %T Hierarchical Conditional Random Fields for Outlier Detection: An Application to Detecting Epileptogenic Cortical Malformations %A Bilal Ahmed %A Thomas Thesen %A Karen Blackmon %A Yijun Zhao %A Orrin Devinsky %A Ruben Kuzniecky %A Carla Brodley %B Proceedings of the 31st International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2014 %E Eric P. Xing %E Tony Jebara %F pmlr-v32-ahmed14 %I PMLR %P 1080--1088 %U https://proceedings.mlr.press/v32/ahmed14.html %V 32 %N 2 %X We cast the problem of detecting and isolating regions of abnormal cortical tissue in the MRIs of epilepsy patients in an image segmentation framework. Employing a multiscale approach we divide the surface images into segments of different sizes and then classify each segment as being an outlier, by comparing it to the same region across controls. The final classification is obtained by fusing the outlier probabilities obtained at multiple scales using a tree-structured hierarchical conditional random field (HCRF). The proposed method correctly detects abnormal regions in 90% of patients whose abnormality was detected via routine visual inspection of their clinical MRI. More importantly, it detects abnormalities in 80% of patients whose abnormality escaped visual inspection by expert radiologists.
RIS
TY - CPAPER TI - Hierarchical Conditional Random Fields for Outlier Detection: An Application to Detecting Epileptogenic Cortical Malformations AU - Bilal Ahmed AU - Thomas Thesen AU - Karen Blackmon AU - Yijun Zhao AU - Orrin Devinsky AU - Ruben Kuzniecky AU - Carla Brodley BT - Proceedings of the 31st International Conference on Machine Learning DA - 2014/06/18 ED - Eric P. Xing ED - Tony Jebara ID - pmlr-v32-ahmed14 PB - PMLR DP - Proceedings of Machine Learning Research VL - 32 IS - 2 SP - 1080 EP - 1088 L1 - http://proceedings.mlr.press/v32/ahmed14.pdf UR - https://proceedings.mlr.press/v32/ahmed14.html AB - We cast the problem of detecting and isolating regions of abnormal cortical tissue in the MRIs of epilepsy patients in an image segmentation framework. Employing a multiscale approach we divide the surface images into segments of different sizes and then classify each segment as being an outlier, by comparing it to the same region across controls. The final classification is obtained by fusing the outlier probabilities obtained at multiple scales using a tree-structured hierarchical conditional random field (HCRF). The proposed method correctly detects abnormal regions in 90% of patients whose abnormality was detected via routine visual inspection of their clinical MRI. More importantly, it detects abnormalities in 80% of patients whose abnormality escaped visual inspection by expert radiologists. ER -
APA
Ahmed, B., Thesen, T., Blackmon, K., Zhao, Y., Devinsky, O., Kuzniecky, R. & Brodley, C.. (2014). Hierarchical Conditional Random Fields for Outlier Detection: An Application to Detecting Epileptogenic Cortical Malformations. Proceedings of the 31st International Conference on Machine Learning, in Proceedings of Machine Learning Research 32(2):1080-1088 Available from https://proceedings.mlr.press/v32/ahmed14.html.

Related Material