Speckle and Shadows: Ultrasound-specific Physics-based Data Augmentation for Kidney Segmentation

Rohit Singla, Cailin Ringstrom, Ricky Hu, Victoria Lessoway, Janice Reid, Robert Rohling, Christophe Nguan
Proceedings of The 5th International Conference on Medical Imaging with Deep Learning, PMLR 172:1139-1148, 2022.

Abstract

Techniques for data augmentation are widely employed to avoid overfitting, improve generalizability and overcome data scarcity. This data-oriented approach frequently uses domain-agnostic approaches such as geometric transformations, colour space transformations, and generative adversarial networks. However, utilsing domain-specific characteristics in augmentations may result in additional invariances or improved robustness. We present several augmentation techniques for ultrasound: zoom, time-gain compensation, artificial shadowing, and speckle parameter maps. Zoom and time-gain compensation mimic traditional image quality parameters. For shadowing, we characterize acoustic shadows within abdominal ultrasound images and provide a method for incorporating artificial shadows into existing images. Finally, we transform B-mode ultrasound images into Nakagami-based speckle parameter maps to describe spatial structures that are not visible in conventional B-mode. The augmentations are evaluated by training a fully supervised network and a contrastive learning network for multi-class intra-organ semantic segmentation. Our preliminary results reflect the difficulties of creating augmentations as well as the limitations posed by acoustic shadowing.

Cite this Paper

BibTeX
@InProceedings{pmlr-v172-singla22a, title = {Speckle and Shadows: Ultrasound-specific Physics-based Data Augmentation for Kidney Segmentation}, author = {Singla, Rohit and Ringstrom, Cailin and Hu, Ricky and Lessoway, Victoria and Reid, Janice and Rohling, Robert and Nguan, Christophe}, booktitle = {Proceedings of The 5th International Conference on Medical Imaging with Deep Learning}, pages = {1139--1148}, 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/singla22a/singla22a.pdf}, url = {https://proceedings.mlr.press/v172/singla22a.html}, abstract = {Techniques for data augmentation are widely employed to avoid overfitting, improve generalizability and overcome data scarcity. This data-oriented approach frequently uses domain-agnostic approaches such as geometric transformations, colour space transformations, and generative adversarial networks. However, utilsing domain-specific characteristics in augmentations may result in additional invariances or improved robustness. We present several augmentation techniques for ultrasound: zoom, time-gain compensation, artificial shadowing, and speckle parameter maps. Zoom and time-gain compensation mimic traditional image quality parameters. For shadowing, we characterize acoustic shadows within abdominal ultrasound images and provide a method for incorporating artificial shadows into existing images. Finally, we transform B-mode ultrasound images into Nakagami-based speckle parameter maps to describe spatial structures that are not visible in conventional B-mode. The augmentations are evaluated by training a fully supervised network and a contrastive learning network for multi-class intra-organ semantic segmentation. Our preliminary results reflect the difficulties of creating augmentations as well as the limitations posed by acoustic shadowing.} }
Endnote
%0 Conference Paper %T Speckle and Shadows: Ultrasound-specific Physics-based Data Augmentation for Kidney Segmentation %A Rohit Singla %A Cailin Ringstrom %A Ricky Hu %A Victoria Lessoway %A Janice Reid %A Robert Rohling %A Christophe Nguan %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-singla22a %I PMLR %P 1139--1148 %U https://proceedings.mlr.press/v172/singla22a.html %V 172 %X Techniques for data augmentation are widely employed to avoid overfitting, improve generalizability and overcome data scarcity. This data-oriented approach frequently uses domain-agnostic approaches such as geometric transformations, colour space transformations, and generative adversarial networks. However, utilsing domain-specific characteristics in augmentations may result in additional invariances or improved robustness. We present several augmentation techniques for ultrasound: zoom, time-gain compensation, artificial shadowing, and speckle parameter maps. Zoom and time-gain compensation mimic traditional image quality parameters. For shadowing, we characterize acoustic shadows within abdominal ultrasound images and provide a method for incorporating artificial shadows into existing images. Finally, we transform B-mode ultrasound images into Nakagami-based speckle parameter maps to describe spatial structures that are not visible in conventional B-mode. The augmentations are evaluated by training a fully supervised network and a contrastive learning network for multi-class intra-organ semantic segmentation. Our preliminary results reflect the difficulties of creating augmentations as well as the limitations posed by acoustic shadowing.
APA
Singla, R., Ringstrom, C., Hu, R., Lessoway, V., Reid, J., Rohling, R. & Nguan, C.. (2022). Speckle and Shadows: Ultrasound-specific Physics-based Data Augmentation for Kidney Segmentation. Proceedings of The 5th International Conference on Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 172:1139-1148 Available from https://proceedings.mlr.press/v172/singla22a.html.

Related Material