Learning Strategies for Contrast-agnostic Segmentation via SynthSeg for Infant MRI data

Ziyao Shang, Md Asadullah Turja, Eric Feczko, Audrey Houghton, Amanda Rueter, Lucille A Moore, Kathy Snider, Timothy Hendrickson, Paul Reiners, Sally Stoyell, Omid Kardan, Monica Rosenberg, Jed T Elison, Damien A Fair, Martin A Styner
Proceedings of The 5th International Conference on Medical Imaging with Deep Learning, PMLR 172:1075-1084, 2022.

Abstract

Longitudinal studies of infants’ brains are essential for research and clinical detection of neurodevelopmental disorders. However, for infant brain MRI scans, effective deep learning-based segmentation frameworks exist only within small age intervals due to the large image intensity and contrast changes that take place in the early postnatal stages of development. However, using different segmentation frameworks or models at different age intervals within the same longitudinal data set would cause segmentation inconsistencies and age-specific biases. Thus, an age-agnostic segmentation model for infants’ brains is needed. In this paper, we present “Infant-SynthSeg“, an extension of the contrast-agnostic SynthSeg segmentation framework applicable to MRI data of infants at ages within the first year of life. Our work mainly focuses on extending learning strategies related to synthetic data generation and augmentation, with the aim of creating a method that employs training data capturing features unique to infants’ brains during this early-stage development. Comparison across different learning strategy settings, as well as a more-traditional contrast-aware deep learning model (nnU-net) are presented. Our experiments show that our trained Infant-SynthSeg models show consistently high segmentation performance on MRI scans of infant brains throughout the first year of life. Furthermore, as the model is trained on ground truth labels at different ages, even labels that are not present at certain ages (such as cerebellar white matter at 1 month) can be appropriately segmented via Infant-SynthSeg across the whole age range. Finally, while Infant-SynthSeg shows consistent segmentation performance across the first year of life, it is outperformed by age-specific deep learning models trained for a specific narrow age range.

Cite this Paper

BibTeX
@InProceedings{pmlr-v172-shang22a, title = {Learning Strategies for Contrast-agnostic Segmentation via SynthSeg for Infant MRI data}, author = {Shang, Ziyao and Turja, Md Asadullah and Feczko, Eric and Houghton, Audrey and Rueter, Amanda and A Moore, Lucille and Snider, Kathy and Hendrickson, Timothy and Reiners, Paul and Stoyell, Sally and Kardan, Omid and Rosenberg, Monica and Elison, Jed T and Fair, Damien A and Styner, Martin A}, booktitle = {Proceedings of The 5th International Conference on Medical Imaging with Deep Learning}, pages = {1075--1084}, 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/shang22a/shang22a.pdf}, url = {https://proceedings.mlr.press/v172/shang22a.html}, abstract = {Longitudinal studies of infants’ brains are essential for research and clinical detection of neurodevelopmental disorders. However, for infant brain MRI scans, effective deep learning-based segmentation frameworks exist only within small age intervals due to the large image intensity and contrast changes that take place in the early postnatal stages of development. However, using different segmentation frameworks or models at different age intervals within the same longitudinal data set would cause segmentation inconsistencies and age-specific biases. Thus, an age-agnostic segmentation model for infants’ brains is needed. In this paper, we present “Infant-SynthSeg“, an extension of the contrast-agnostic SynthSeg segmentation framework applicable to MRI data of infants at ages within the first year of life. Our work mainly focuses on extending learning strategies related to synthetic data generation and augmentation, with the aim of creating a method that employs training data capturing features unique to infants’ brains during this early-stage development. Comparison across different learning strategy settings, as well as a more-traditional contrast-aware deep learning model (nnU-net) are presented. Our experiments show that our trained Infant-SynthSeg models show consistently high segmentation performance on MRI scans of infant brains throughout the first year of life. Furthermore, as the model is trained on ground truth labels at different ages, even labels that are not present at certain ages (such as cerebellar white matter at 1 month) can be appropriately segmented via Infant-SynthSeg across the whole age range. Finally, while Infant-SynthSeg shows consistent segmentation performance across the first year of life, it is outperformed by age-specific deep learning models trained for a specific narrow age range.} }
Endnote
%0 Conference Paper %T Learning Strategies for Contrast-agnostic Segmentation via SynthSeg for Infant MRI data %A Ziyao Shang %A Md Asadullah Turja %A Eric Feczko %A Audrey Houghton %A Amanda Rueter %A Lucille A Moore %A Kathy Snider %A Timothy Hendrickson %A Paul Reiners %A Sally Stoyell %A Omid Kardan %A Monica Rosenberg %A Jed T Elison %A Damien A Fair %A Martin A Styner %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-shang22a %I PMLR %P 1075--1084 %U https://proceedings.mlr.press/v172/shang22a.html %V 172 %X Longitudinal studies of infants’ brains are essential for research and clinical detection of neurodevelopmental disorders. However, for infant brain MRI scans, effective deep learning-based segmentation frameworks exist only within small age intervals due to the large image intensity and contrast changes that take place in the early postnatal stages of development. However, using different segmentation frameworks or models at different age intervals within the same longitudinal data set would cause segmentation inconsistencies and age-specific biases. Thus, an age-agnostic segmentation model for infants’ brains is needed. In this paper, we present “Infant-SynthSeg“, an extension of the contrast-agnostic SynthSeg segmentation framework applicable to MRI data of infants at ages within the first year of life. Our work mainly focuses on extending learning strategies related to synthetic data generation and augmentation, with the aim of creating a method that employs training data capturing features unique to infants’ brains during this early-stage development. Comparison across different learning strategy settings, as well as a more-traditional contrast-aware deep learning model (nnU-net) are presented. Our experiments show that our trained Infant-SynthSeg models show consistently high segmentation performance on MRI scans of infant brains throughout the first year of life. Furthermore, as the model is trained on ground truth labels at different ages, even labels that are not present at certain ages (such as cerebellar white matter at 1 month) can be appropriately segmented via Infant-SynthSeg across the whole age range. Finally, while Infant-SynthSeg shows consistent segmentation performance across the first year of life, it is outperformed by age-specific deep learning models trained for a specific narrow age range.
APA
Shang, Z., Turja, M.A., Feczko, E., Houghton, A., Rueter, A., A Moore, L., Snider, K., Hendrickson, T., Reiners, P., Stoyell, S., Kardan, O., Rosenberg, M., Elison, J.T., Fair, D.A. & Styner, M.A.. (2022). Learning Strategies for Contrast-agnostic Segmentation via SynthSeg for Infant MRI data. Proceedings of The 5th International Conference on Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 172:1075-1084 Available from https://proceedings.mlr.press/v172/shang22a.html.

Related Material