Thyroid Cancer Malignancy Prediction From Whole Slide Cytopathology Images

David Dov, Shahar Z. Kovalsky, Jonathan Cohen, Danielle Elliott Range, Ricardo Henao, Lawrence Carin
Proceedings of the 4th Machine Learning for Healthcare Conference, PMLR 106:553-570, 2019.

Abstract

We consider preoperative prediction of thyroid cancer based on ultra-high-resolution whole-slide cytopathology images. Inspired by how human experts perform diagnosis, our approach first identifies and classifies diagnostic image regions containing informative thyroid cells, which only comprise a tiny fraction of the entire image. These local estimates are then aggregated into a single prediction of thyroid malignancy. Several unique characteristics of thyroid cytopathology guide our deep-learning-based approach. While our method is closely related to multiple-instance learning, it deviates from these methods by using a supervised procedure to extract diagnostically relevant regions. Moreover, we propose to simultaneously predict thyroid malignancy, as well as a diagnostic score assigned by a human expert, which further allows us to devise an improved training strategy. Experimental results show that the proposed algorithm achieves performance comparable to human experts, and demonstrate the potential of using the algorithm for screening and as an assistive tool for the improved diagnosis of indeterminate cases.

Cite this Paper

BibTeX
@InProceedings{pmlr-v106-dov19a, title = {Thyroid Cancer Malignancy Prediction From Whole Slide Cytopathology Images}, author = {Dov, David and Kovalsky, Shahar Z. and Cohen, Jonathan and Range, Danielle Elliott and Henao, Ricardo and Carin, Lawrence}, booktitle = {Proceedings of the 4th Machine Learning for Healthcare Conference}, pages = {553--570}, year = {2019}, editor = {Doshi-Velez, Finale and Fackler, Jim and Jung, Ken and Kale, David and Ranganath, Rajesh and Wallace, Byron and Wiens, Jenna}, volume = {106}, series = {Proceedings of Machine Learning Research}, month = {09--10 Aug}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v106/dov19a/dov19a.pdf}, url = {https://proceedings.mlr.press/v106/dov19a.html}, abstract = {We consider preoperative prediction of thyroid cancer based on ultra-high-resolution whole-slide cytopathology images. Inspired by how human experts perform diagnosis, our approach first identifies and classifies diagnostic image regions containing informative thyroid cells, which only comprise a tiny fraction of the entire image. These local estimates are then aggregated into a single prediction of thyroid malignancy. Several unique characteristics of thyroid cytopathology guide our deep-learning-based approach. While our method is closely related to multiple-instance learning, it deviates from these methods by using a supervised procedure to extract diagnostically relevant regions. Moreover, we propose to simultaneously predict thyroid malignancy, as well as a diagnostic score assigned by a human expert, which further allows us to devise an improved training strategy. Experimental results show that the proposed algorithm achieves performance comparable to human experts, and demonstrate the potential of using the algorithm for screening and as an assistive tool for the improved diagnosis of indeterminate cases.} }
Endnote
%0 Conference Paper %T Thyroid Cancer Malignancy Prediction From Whole Slide Cytopathology Images %A David Dov %A Shahar Z. Kovalsky %A Jonathan Cohen %A Danielle Elliott Range %A Ricardo Henao %A Lawrence Carin %B Proceedings of the 4th Machine Learning for Healthcare Conference %C Proceedings of Machine Learning Research %D 2019 %E Finale Doshi-Velez %E Jim Fackler %E Ken Jung %E David Kale %E Rajesh Ranganath %E Byron Wallace %E Jenna Wiens %F pmlr-v106-dov19a %I PMLR %P 553--570 %U https://proceedings.mlr.press/v106/dov19a.html %V 106 %X We consider preoperative prediction of thyroid cancer based on ultra-high-resolution whole-slide cytopathology images. Inspired by how human experts perform diagnosis, our approach first identifies and classifies diagnostic image regions containing informative thyroid cells, which only comprise a tiny fraction of the entire image. These local estimates are then aggregated into a single prediction of thyroid malignancy. Several unique characteristics of thyroid cytopathology guide our deep-learning-based approach. While our method is closely related to multiple-instance learning, it deviates from these methods by using a supervised procedure to extract diagnostically relevant regions. Moreover, we propose to simultaneously predict thyroid malignancy, as well as a diagnostic score assigned by a human expert, which further allows us to devise an improved training strategy. Experimental results show that the proposed algorithm achieves performance comparable to human experts, and demonstrate the potential of using the algorithm for screening and as an assistive tool for the improved diagnosis of indeterminate cases.
APA
Dov, D., Kovalsky, S.Z., Cohen, J., Range, D.E., Henao, R. & Carin, L.. (2019). Thyroid Cancer Malignancy Prediction From Whole Slide Cytopathology Images. Proceedings of the 4th Machine Learning for Healthcare Conference, in Proceedings of Machine Learning Research 106:553-570 Available from https://proceedings.mlr.press/v106/dov19a.html.

Related Material