RadBERT-CL: Factually-Aware Contrastive Learning For Radiology Report Classification

Ajay Jaiswal, Liyan Tang, Meheli Ghosh, Justin F. Rousseau, Yifan Peng, Ying Ding
Proceedings of Machine Learning for Health, PMLR 158:196-208, 2021.

Abstract

Radiology reports are unstructured and contain the imaging findings and corresponding diagnoses transcribed by radiologists which include clinical facts and negated and/or uncertain statements. Extracting pathologic findings and diagnoses from radiology reports is important for quality control, population health, and monitoring of disease progress. Existing works, primarily rely either on rule-based systems or transformer-based pre-trained model fine-tuning, but could not take the factual and uncertain information into consideration, and therefore generate false positive outputs. In this work, we introduce three sedulous augmentation techniques which retain factual and critical information while generating augmentations for contrastive learning. We introduce RadBERT-CL, which fuses these information into BlueBert via a self-supervised contrastive loss. Our experiments on MIMIC-CXR show superior performance of RadBERT-CL on fine-tuning for multi-class, multi-label report classification. We illustrate that when few labeled data are available, RadBERT-CL outperforms conventional SOTA transformers (BERT/BlueBert) by significantly larger margins (6-11%). We also show that the representations learned by RadBERT-CL can capture critical medical information in the latent space.

Cite this Paper


BibTeX
@InProceedings{pmlr-v158-jaiswal21a, title = {RadBERT-CL: Factually-Aware Contrastive Learning For Radiology Report Classification}, author = {Jaiswal, Ajay and Tang, Liyan and Ghosh, Meheli and Rousseau, Justin F. and Peng, Yifan and Ding, Ying}, booktitle = {Proceedings of Machine Learning for Health}, pages = {196--208}, year = {2021}, editor = {Roy, Subhrajit and Pfohl, Stephen and Rocheteau, Emma and Tadesse, Girmaw Abebe and Oala, Luis and Falck, Fabian and Zhou, Yuyin and Shen, Liyue and Zamzmi, Ghada and Mugambi, Purity and Zirikly, Ayah and McDermott, Matthew B. A. and Alsentzer, Emily}, volume = {158}, series = {Proceedings of Machine Learning Research}, month = {04 Dec}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v158/jaiswal21a/jaiswal21a.pdf}, url = {https://proceedings.mlr.press/v158/jaiswal21a.html}, abstract = {Radiology reports are unstructured and contain the imaging findings and corresponding diagnoses transcribed by radiologists which include clinical facts and negated and/or uncertain statements. Extracting pathologic findings and diagnoses from radiology reports is important for quality control, population health, and monitoring of disease progress. Existing works, primarily rely either on rule-based systems or transformer-based pre-trained model fine-tuning, but could not take the factual and uncertain information into consideration, and therefore generate false positive outputs. In this work, we introduce three sedulous augmentation techniques which retain factual and critical information while generating augmentations for contrastive learning. We introduce RadBERT-CL, which fuses these information into BlueBert via a self-supervised contrastive loss. Our experiments on MIMIC-CXR show superior performance of RadBERT-CL on fine-tuning for multi-class, multi-label report classification. We illustrate that when few labeled data are available, RadBERT-CL outperforms conventional SOTA transformers (BERT/BlueBert) by significantly larger margins (6-11%). We also show that the representations learned by RadBERT-CL can capture critical medical information in the latent space.} }
Endnote
%0 Conference Paper %T RadBERT-CL: Factually-Aware Contrastive Learning For Radiology Report Classification %A Ajay Jaiswal %A Liyan Tang %A Meheli Ghosh %A Justin F. Rousseau %A Yifan Peng %A Ying Ding %B Proceedings of Machine Learning for Health %C Proceedings of Machine Learning Research %D 2021 %E Subhrajit Roy %E Stephen Pfohl %E Emma Rocheteau %E Girmaw Abebe Tadesse %E Luis Oala %E Fabian Falck %E Yuyin Zhou %E Liyue Shen %E Ghada Zamzmi %E Purity Mugambi %E Ayah Zirikly %E Matthew B. A. McDermott %E Emily Alsentzer %F pmlr-v158-jaiswal21a %I PMLR %P 196--208 %U https://proceedings.mlr.press/v158/jaiswal21a.html %V 158 %X Radiology reports are unstructured and contain the imaging findings and corresponding diagnoses transcribed by radiologists which include clinical facts and negated and/or uncertain statements. Extracting pathologic findings and diagnoses from radiology reports is important for quality control, population health, and monitoring of disease progress. Existing works, primarily rely either on rule-based systems or transformer-based pre-trained model fine-tuning, but could not take the factual and uncertain information into consideration, and therefore generate false positive outputs. In this work, we introduce three sedulous augmentation techniques which retain factual and critical information while generating augmentations for contrastive learning. We introduce RadBERT-CL, which fuses these information into BlueBert via a self-supervised contrastive loss. Our experiments on MIMIC-CXR show superior performance of RadBERT-CL on fine-tuning for multi-class, multi-label report classification. We illustrate that when few labeled data are available, RadBERT-CL outperforms conventional SOTA transformers (BERT/BlueBert) by significantly larger margins (6-11%). We also show that the representations learned by RadBERT-CL can capture critical medical information in the latent space.
APA
Jaiswal, A., Tang, L., Ghosh, M., Rousseau, J.F., Peng, Y. & Ding, Y.. (2021). RadBERT-CL: Factually-Aware Contrastive Learning For Radiology Report Classification. Proceedings of Machine Learning for Health, in Proceedings of Machine Learning Research 158:196-208 Available from https://proceedings.mlr.press/v158/jaiswal21a.html.

Related Material