Explainability Guided COVID-19 Detection in CT Scans

Ameen Ali, Tal Shaharabany, Lior Wolf
Proceedings of The 5th International Conference on Medical Imaging with Deep Learning, PMLR 172:5-21, 2022.

Abstract

Radiological examination of chest CT is an effective method for screening COVID-19 cases. In this work, we overcome three challenges in the automation of this process: (i) the limited number of supervised positive cases, (ii) the lack of region-based supervision, and (iii) variability across acquisition sites. These challenges are met by incorporating a recent augmentation solution called SnapMix, a novel explainability-driven contrastive loss for patch embedding, and by performing test-time augmentation that masks out the most relevant patches in order to analyse the prediction stability. The three techniques are complementary and are all based on utilizing the heatmaps produced by the Class Activation Mapping (CAM) explainability method. State-of-the-art performance is obtained on three different datasets for COVID detection in CT scans.

Cite this Paper

BibTeX
@InProceedings{pmlr-v172-ali22a, title = {Explainability Guided COVID-19 Detection in CT Scans}, author = {Ali, Ameen and Shaharabany, Tal and Wolf, Lior}, booktitle = {Proceedings of The 5th International Conference on Medical Imaging with Deep Learning}, pages = {5--21}, 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/ali22a/ali22a.pdf}, url = {https://proceedings.mlr.press/v172/ali22a.html}, abstract = {Radiological examination of chest CT is an effective method for screening COVID-19 cases. In this work, we overcome three challenges in the automation of this process: (i) the limited number of supervised positive cases, (ii) the lack of region-based supervision, and (iii) variability across acquisition sites. These challenges are met by incorporating a recent augmentation solution called SnapMix, a novel explainability-driven contrastive loss for patch embedding, and by performing test-time augmentation that masks out the most relevant patches in order to analyse the prediction stability. The three techniques are complementary and are all based on utilizing the heatmaps produced by the Class Activation Mapping (CAM) explainability method. State-of-the-art performance is obtained on three different datasets for COVID detection in CT scans.} }
Endnote
%0 Conference Paper %T Explainability Guided COVID-19 Detection in CT Scans %A Ameen Ali %A Tal Shaharabany %A Lior Wolf %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-ali22a %I PMLR %P 5--21 %U https://proceedings.mlr.press/v172/ali22a.html %V 172 %X Radiological examination of chest CT is an effective method for screening COVID-19 cases. In this work, we overcome three challenges in the automation of this process: (i) the limited number of supervised positive cases, (ii) the lack of region-based supervision, and (iii) variability across acquisition sites. These challenges are met by incorporating a recent augmentation solution called SnapMix, a novel explainability-driven contrastive loss for patch embedding, and by performing test-time augmentation that masks out the most relevant patches in order to analyse the prediction stability. The three techniques are complementary and are all based on utilizing the heatmaps produced by the Class Activation Mapping (CAM) explainability method. State-of-the-art performance is obtained on three different datasets for COVID detection in CT scans.
APA
Ali, A., Shaharabany, T. & Wolf, L.. (2022). Explainability Guided COVID-19 Detection in CT Scans. Proceedings of The 5th International Conference on Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 172:5-21 Available from https://proceedings.mlr.press/v172/ali22a.html.

Related Material