Evaluating and interpreting caption prediction for histopathology images

Renyu Zhang, Christopher Weber, Robert Grossman, Aly A. Khan
Proceedings of the 5th Machine Learning for Healthcare Conference, PMLR 126:418-435, 2020.

Abstract

The automatic generation of captions from medical images can provide for an efficient way to annotate histopathology images with natural language descriptions. Such large-scale annotation of medical images may help facilitate image retrieval tasks and standardize clinical ontologies. In this work, we focus on developing and methodically evaluating a new caption generation framework for histopathology whole-slide images. We introduce PathCap, a deep learning multi-scale framework, to predict captions from histopathology images using multi-scale views of whole-slide images. We demonstrate that our framework outperforms a standard baseline caption model on a diverse set of human tissues and provides interpretable contextual cues for understanding predicted captions. Finally, we draw attention to a novel dataset of histopathology images with captions from the Genotype-Tissue Expression (GTEx) project, providing a valuable dataset for the machine learning and healthcare community to benchmark future caption prediction and interpretation methods.

Cite this Paper


BibTeX
@InProceedings{pmlr-v126-zhang20b, title = {Evaluating and interpreting caption prediction for histopathology images}, author = {Zhang, Renyu and Weber, Christopher and Grossman, Robert and Khan, Aly A.}, booktitle = {Proceedings of the 5th Machine Learning for Healthcare Conference}, pages = {418--435}, year = {2020}, editor = {Doshi-Velez, Finale and Fackler, Jim and Jung, Ken and Kale, David and Ranganath, Rajesh and Wallace, Byron and Wiens, Jenna}, volume = {126}, series = {Proceedings of Machine Learning Research}, month = {07--08 Aug}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v126/zhang20b/zhang20b.pdf}, url = {https://proceedings.mlr.press/v126/zhang20b.html}, abstract = {The automatic generation of captions from medical images can provide for an efficient way to annotate histopathology images with natural language descriptions. Such large-scale annotation of medical images may help facilitate image retrieval tasks and standardize clinical ontologies. In this work, we focus on developing and methodically evaluating a new caption generation framework for histopathology whole-slide images. We introduce PathCap, a deep learning multi-scale framework, to predict captions from histopathology images using multi-scale views of whole-slide images. We demonstrate that our framework outperforms a standard baseline caption model on a diverse set of human tissues and provides interpretable contextual cues for understanding predicted captions. Finally, we draw attention to a novel dataset of histopathology images with captions from the Genotype-Tissue Expression (GTEx) project, providing a valuable dataset for the machine learning and healthcare community to benchmark future caption prediction and interpretation methods.} }
Endnote
%0 Conference Paper %T Evaluating and interpreting caption prediction for histopathology images %A Renyu Zhang %A Christopher Weber %A Robert Grossman %A Aly A. Khan %B Proceedings of the 5th Machine Learning for Healthcare Conference %C Proceedings of Machine Learning Research %D 2020 %E Finale Doshi-Velez %E Jim Fackler %E Ken Jung %E David Kale %E Rajesh Ranganath %E Byron Wallace %E Jenna Wiens %F pmlr-v126-zhang20b %I PMLR %P 418--435 %U https://proceedings.mlr.press/v126/zhang20b.html %V 126 %X The automatic generation of captions from medical images can provide for an efficient way to annotate histopathology images with natural language descriptions. Such large-scale annotation of medical images may help facilitate image retrieval tasks and standardize clinical ontologies. In this work, we focus on developing and methodically evaluating a new caption generation framework for histopathology whole-slide images. We introduce PathCap, a deep learning multi-scale framework, to predict captions from histopathology images using multi-scale views of whole-slide images. We demonstrate that our framework outperforms a standard baseline caption model on a diverse set of human tissues and provides interpretable contextual cues for understanding predicted captions. Finally, we draw attention to a novel dataset of histopathology images with captions from the Genotype-Tissue Expression (GTEx) project, providing a valuable dataset for the machine learning and healthcare community to benchmark future caption prediction and interpretation methods.
APA
Zhang, R., Weber, C., Grossman, R. & Khan, A.A.. (2020). Evaluating and interpreting caption prediction for histopathology images. Proceedings of the 5th Machine Learning for Healthcare Conference, in Proceedings of Machine Learning Research 126:418-435 Available from https://proceedings.mlr.press/v126/zhang20b.html.

Related Material