TY - JOUR
T1 - Efficient semi-supervised semantic segmentation of electron microscopy cancer images with sparse annotations
AU - Pagano, Lucas
AU - Thibault, Guillaume
AU - Bousselham, Walid
AU - Riesterer, Jessica L.
AU - Song, Xubo
AU - Gray, Joe
N1 - Publisher Copyright:
Copyright © 2023 Pagano, Thibault, Bousselham, Riesterer, Song and Gray.
PY - 2023
Y1 - 2023
N2 - Electron microscopy (EM) enables imaging at a resolution of nanometers and can shed light on how cancer evolves to develop resistance to therapy. Acquiring these images has become a routine task.However, analyzing them is now a bottleneck, as manual structure identification is very time-consuming and can take up to several months for a single sample. Deep learning approaches offer a suitable solution to speed up the analysis. In this work, we present a study of several state-of-the-art deep learning models for the task of segmenting nuclei and nucleoli in volumes from tumor biopsies. We compared previous results obtained with the ResUNet architecture to the more recent UNet++, FracTALResNet, SenFormer, and CEECNet models. In addition, we explored the utilization of unlabeled images through semi-supervised learning with Cross Pseudo Supervision. We have trained and evaluated all of the models on sparse manual labels from three fully annotated in-house datasets that we have made available on demand, demonstrating improvements in terms of 3D Dice score. From the analysis of these results, we drew conclusions on the relative gains of using more complex models, and semi-supervised learning as well as the next steps for the mitigation of the manual segmentation bottleneck.
AB - Electron microscopy (EM) enables imaging at a resolution of nanometers and can shed light on how cancer evolves to develop resistance to therapy. Acquiring these images has become a routine task.However, analyzing them is now a bottleneck, as manual structure identification is very time-consuming and can take up to several months for a single sample. Deep learning approaches offer a suitable solution to speed up the analysis. In this work, we present a study of several state-of-the-art deep learning models for the task of segmenting nuclei and nucleoli in volumes from tumor biopsies. We compared previous results obtained with the ResUNet architecture to the more recent UNet++, FracTALResNet, SenFormer, and CEECNet models. In addition, we explored the utilization of unlabeled images through semi-supervised learning with Cross Pseudo Supervision. We have trained and evaluated all of the models on sparse manual labels from three fully annotated in-house datasets that we have made available on demand, demonstrating improvements in terms of 3D Dice score. From the analysis of these results, we drew conclusions on the relative gains of using more complex models, and semi-supervised learning as well as the next steps for the mitigation of the manual segmentation bottleneck.
KW - deep learning
KW - electron microscopy
KW - semantic segmentation
KW - semi-supervised
KW - sparse labels
KW - vEM
UR - http://www.scopus.com/inward/record.url?scp=85180884074&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85180884074&partnerID=8YFLogxK
U2 - 10.3389/fbinf.2023.1308707
DO - 10.3389/fbinf.2023.1308707
M3 - Article
AN - SCOPUS:85180884074
SN - 2673-7647
VL - 3
JO - Frontiers in Bioinformatics
JF - Frontiers in Bioinformatics
M1 - 1308707
ER -