SlideShare a Scribd company logo
1 of 15
Hierarchical Consistency Regularized Mean Teacher for
Semi-supervised 3D Left Atrium Segmentation
Shumeng Li*1,2, Ziyuan Zhao*1,2 , Kaixin Xu2, Zeng Zeng2, Cuntai Guan1
Presenter: Zhao Ziyuan
1 Nanyang Technological University 2 Institute for Infocomm Research, A*STAR
Background
Natural Images Medical Images
Need different pre-processing methods Not many classes
To segment certain specific structures or tissues
● Image segmentation -> assign a label to every pixel
Problem Statement
Challenge for medical images:
● Requires manual identification by
experts
● Annotations are difficult to obtain
Semi-supervised
learning
Full-supervised
learning
Data
Data
Labeled
data
Labeled
data
Unlabeled
data
Improve the segmentation
performance with the
unlabeled data
Proposed Method
- Overview
16
128
64
32
16
32
64
128
256
Lc0 Lc1 Lc2 Lc3
Lunsu
p
Ls0
Ls1
Ls2
Ls3
Lsu
p
Noise 𝞷
Noise 𝞷’
Teacher
Network
concatenate
downsample conv
upsample deconv
upsample, conv 1×1, softmax
Ground
Truth
Upsample
× 2
Upsample
× 4
Upsample
× 8
Prediction
Input
16
128
64
32
16
32
64
128
256
Student
Network
Upsample
× 2
Upsample
× 4
Upsample
× 8
EMA
(a) Multi-scale deeply supervised network
(b) Hierarchical consistency regularized mean teacher
Proposed Method (a)
- Multi-scale deeply supervised network
Ls0
Ls1
Ls2
Ls3
Lsu
p
concatenate
downsample conv
upsample deconv
upsample, conv 1×1, softmax
Ground
Truth
Prediction
16
128
64
32
16
32
64
128
256
V-Net
Upsample
× 2
Upsample
× 4
Upsample
× 8
● Hierarchical supervised (HS) component
○ add the auxiliary layers after each block in the decoding stage of V-Net
■ an upsampling layer, a 1x1x1 convolution, and a softmax layer
○ multiple prediction results form multi-scale deep supervision
● Supervised loss
○ dice loss + cross-entropy loss
Lsu
p
Lunsu
p
Proposed Method (b)
- Basic mean teacher (MT)
16
128
64
32
16
32
64
128
256
Noise 𝞷
Noise 𝞷’
Teacher
Network
Ground
Truth
Prediction
Input
16
128
64
32
16
32
64
128
256
Student
Network
EMA
concatenate
downsample conv
upsample deconv
upsample, conv 1×1, softmax
● Semi-supervised learning
○ objective function: supervised loss Lsup + unsupervised consistency loss Lunsup
the teacher network weights are
updated with the exponential
moving average (EMA) of the
student network weights
Labeled &
Unlabeled
Labeled
Tarvainen, A., & Valpola, H. (2017). Mean teachers are better role
models: Weight-averaged consistency targets improve semi-
supervised deep learning results. arXiv preprint arXiv:1703.01780.
Proposed Method (b)
- Hierarchical consistency regularized MT
16
128
64
32
16
32
64
128
256
Lc0 Lc1 Lc2 Lc3
Lunsu
p
Noise 𝞷
Noise 𝞷’
Teacher
Network
concatenate
downsample conv
upsample deconv
upsample, conv 1×1, softmax
Upsample
× 2
Upsample
× 4
Upsample
× 8
Prediction
Input
16
128
64
32
16
32
64
128
256
Student
Network
Upsample
× 2
Upsample
× 4
Upsample
× 8
EMA
● Hierarchical unsupervised (HU) component
○ add the auxiliary layers to both student and teacher networks
○ further encourage the prediction consistent in the hidden feature space
Proposed Method (b)
- Hierarchical consistency regularized MT
16
128
64
32
16
32
64
128
256
Lc0 Lc1 Lc2 Lc3
Lunsu
p
Noise 𝞷
Noise 𝞷’
Teacher
Network
concatenate
downsample conv
upsample deconv
upsample, conv 1×1, softmax
Upsample
× 2
Upsample
× 4
Upsample
× 8
Prediction
Input
16
128
64
32
16
32
64
128
256
Student
Network
Upsample
× 2
Upsample
× 4
Upsample
× 8
EMA
● Consistency loss
○ mean squared error (MSE) loss of multi-scale predictions between the student and
the teacher networks
Proposed Method
- Overview
16
128
64
32
16
32
64
128
256
Lc0 Lc1 Lc2 Lc3
Lunsu
p
Ls0
Ls1
Ls2
Ls3
Lsu
p
Noise 𝞷
Noise 𝞷’
Teacher
Network
concatenate
downsample conv
upsample deconv
upsample, conv 1×1, softmax
Ground
Truth
Upsample
× 2
Upsample
× 4
Upsample
× 8
Prediction
Input
16
128
64
32
16
32
64
128
256
Student
Network
Upsample
× 2
Upsample
× 4
Upsample
× 8
EMA
Labeled &
Unlabeled
Labeled
Dataset
● Data source:
○ provided by 2018 Atrial Segmentation Challenge
● Data size:
○ training Set – 80 3D gadolinium-enhanced MRIs and left atrium cavity labels
■ 16 samples - labeled dataset
■ 64 samples - unlabeled dataset
○ test Set – 20 3D MRIs and left atrium cavity labels
● Pre-processing:
○ all the scans were cropped centering at the heart region and normalized as zero
mean and unit variance
Results
● Our approach achieves the best performance
○ over the state-of-the-art semi-supervised methods in dice, jaccard and 95HD
● Our approach effectively utilizes the unlabeled data
○ compared with full annotation, it obtains comparable segmentation results using only
1/5 labeled data
Table 1: Segmentation results of different approaches
Results
UA-MT SASSNet LG-ER-MT DTC Ours GT
● Example of 2D and 3D results of five semi-supervised segmentation methods, where GT
denotes corresponding ground truth labels
Results
● To evaluate the effectiveness of HS and HU
○ HS = hierarchical supervised component
○ HU = hierarchical unsupervised component
● Introduce HS to V-Net
○ better utilize the labeled data
● Introduce HU + HS to MT
○ both are effective and combining together can achieve better performance
Table 2: Effectiveness of core component in the proposed framework
Conclusion
● The proposed framework based on the mean teacher architecture
○ it encourages the consistency of predictions between the student and the teacher
networks with the same input under different perturbations
● In each iteration, the student network is optimized by multi-scale deep supervision and
hierarchical consistency regularization, concurrently
● Experimental results demonstrate the superiority over the other state-of-the-art SSL
methods
● The proposed method is simple and versatile
○ can be widely applied to other segmentation tasks
Thanks
Zhao_Ziyuan@i2r.a-star.edu.sg

More Related Content

Similar to [EMBC 2021] Hierarchical Consistency Regularized Mean Teacher for Semi-supervised 3D Left Atrium Segmentation

Poster_Reseau_Neurones_Journees_2013
Poster_Reseau_Neurones_Journees_2013Poster_Reseau_Neurones_Journees_2013
Poster_Reseau_Neurones_Journees_2013
Pedro Lopes
 

Similar to [EMBC 2021] Hierarchical Consistency Regularized Mean Teacher for Semi-supervised 3D Left Atrium Segmentation (20)

Noise-robust classification with hypergraph neural network
Noise-robust classification with hypergraph neural networkNoise-robust classification with hypergraph neural network
Noise-robust classification with hypergraph neural network
 
IMAGE CLASSIFICATION USING DIFFERENT CLASSICAL APPROACHES
IMAGE CLASSIFICATION USING DIFFERENT CLASSICAL APPROACHESIMAGE CLASSIFICATION USING DIFFERENT CLASSICAL APPROACHES
IMAGE CLASSIFICATION USING DIFFERENT CLASSICAL APPROACHES
 
PaperReview_ “Few-shot Graph Classification with Contrastive Loss and Meta-cl...
PaperReview_ “Few-shot Graph Classification with Contrastive Loss and Meta-cl...PaperReview_ “Few-shot Graph Classification with Contrastive Loss and Meta-cl...
PaperReview_ “Few-shot Graph Classification with Contrastive Loss and Meta-cl...
 
230727_HB_JointJournalClub.pptx
230727_HB_JointJournalClub.pptx230727_HB_JointJournalClub.pptx
230727_HB_JointJournalClub.pptx
 
An Evolutionary-based Neural Network for Distinguishing between Genuine and P...
An Evolutionary-based Neural Network for Distinguishing between Genuine and P...An Evolutionary-based Neural Network for Distinguishing between Genuine and P...
An Evolutionary-based Neural Network for Distinguishing between Genuine and P...
 
Graph Attention Networks.pptx
Graph Attention Networks.pptxGraph Attention Networks.pptx
Graph Attention Networks.pptx
 
Deep Semi-Supervised Anomaly Detection
Deep Semi-Supervised Anomaly DetectionDeep Semi-Supervised Anomaly Detection
Deep Semi-Supervised Anomaly Detection
 
F017533540
F017533540F017533540
F017533540
 
NS-CUK Journal club: HBKim, Review on "Neural Graph Collaborative Filtering",...
NS-CUK Journal club: HBKim, Review on "Neural Graph Collaborative Filtering",...NS-CUK Journal club: HBKim, Review on "Neural Graph Collaborative Filtering",...
NS-CUK Journal club: HBKim, Review on "Neural Graph Collaborative Filtering",...
 
Artificial Neural Networks Lect3: Neural Network Learning rules
Artificial Neural Networks Lect3: Neural Network Learning rulesArtificial Neural Networks Lect3: Neural Network Learning rules
Artificial Neural Networks Lect3: Neural Network Learning rules
 
Lifelong Learning for Dynamically Expandable Networks
Lifelong Learning for Dynamically Expandable NetworksLifelong Learning for Dynamically Expandable Networks
Lifelong Learning for Dynamically Expandable Networks
 
Deep Semi-Supervised Anomaly Detection
Deep Semi-Supervised Anomaly DetectionDeep Semi-Supervised Anomaly Detection
Deep Semi-Supervised Anomaly Detection
 
A survey on methods and applications of meta-learning with GNNs
A survey on methods and applications of meta-learning with GNNsA survey on methods and applications of meta-learning with GNNs
A survey on methods and applications of meta-learning with GNNs
 
Hidden Layer Leraning Vector Quantizatio
Hidden Layer Leraning Vector Quantizatio Hidden Layer Leraning Vector Quantizatio
Hidden Layer Leraning Vector Quantizatio
 
Hanjun Dai, PhD Student, School of Computational Science and Engineering, Geo...
Hanjun Dai, PhD Student, School of Computational Science and Engineering, Geo...Hanjun Dai, PhD Student, School of Computational Science and Engineering, Geo...
Hanjun Dai, PhD Student, School of Computational Science and Engineering, Geo...
 
Artificial Neural Networks for NIU
Artificial Neural Networks for NIUArtificial Neural Networks for NIU
Artificial Neural Networks for NIU
 
Md2k 0219 shang
Md2k 0219 shangMd2k 0219 shang
Md2k 0219 shang
 
Csc446: Pattern Recognition
Csc446: Pattern Recognition Csc446: Pattern Recognition
Csc446: Pattern Recognition
 
Poster_Reseau_Neurones_Journees_2013
Poster_Reseau_Neurones_Journees_2013Poster_Reseau_Neurones_Journees_2013
Poster_Reseau_Neurones_Journees_2013
 
Deep Learning in Recommender Systems - RecSys Summer School 2017
Deep Learning in Recommender Systems - RecSys Summer School 2017Deep Learning in Recommender Systems - RecSys Summer School 2017
Deep Learning in Recommender Systems - RecSys Summer School 2017
 

More from Ziyuan Zhao

More from Ziyuan Zhao (15)

[IJCAI 2023] SemiGNN-PPI: Self-Ensembling Multi-Graph Neural Network for Effi...
[IJCAI 2023] SemiGNN-PPI: Self-Ensembling Multi-Graph Neural Network for Effi...[IJCAI 2023] SemiGNN-PPI: Self-Ensembling Multi-Graph Neural Network for Effi...
[IJCAI 2023] SemiGNN-PPI: Self-Ensembling Multi-Graph Neural Network for Effi...
 
[IJCAI 2023 - Poster] SemiGNN-PPI: Self-Ensembling Multi-Graph Neural Network...
[IJCAI 2023 - Poster] SemiGNN-PPI: Self-Ensembling Multi-Graph Neural Network...[IJCAI 2023 - Poster] SemiGNN-PPI: Self-Ensembling Multi-Graph Neural Network...
[IJCAI 2023 - Poster] SemiGNN-PPI: Self-Ensembling Multi-Graph Neural Network...
 
[IAIM 2023 - Poster] Label-efficient Generalizable Deep Learning for Medical...
[IAIM 2023 - Poster] Label-efficient Generalizable Deep Learning  for Medical...[IAIM 2023 - Poster] Label-efficient Generalizable Deep Learning  for Medical...
[IAIM 2023 - Poster] Label-efficient Generalizable Deep Learning for Medical...
 
[BMVC 2022 - Spotlight] DA-CIL: Towards Domain Adaptive Class-Incremental 3D ...
[BMVC 2022 - Spotlight] DA-CIL: Towards Domain Adaptive Class-Incremental 3D ...[BMVC 2022 - Spotlight] DA-CIL: Towards Domain Adaptive Class-Incremental 3D ...
[BMVC 2022 - Spotlight] DA-CIL: Towards Domain Adaptive Class-Incremental 3D ...
 
[BMVC 2022] DA-CIL: Towards Domain Adaptive Class-Incremental 3D Object Detec...
[BMVC 2022] DA-CIL: Towards Domain Adaptive Class-Incremental 3D Object Detec...[BMVC 2022] DA-CIL: Towards Domain Adaptive Class-Incremental 3D Object Detec...
[BMVC 2022] DA-CIL: Towards Domain Adaptive Class-Incremental 3D Object Detec...
 
[ICIP 2022] ACT-NET: Asymmetric Co-Teacher Network for Semi-Supervised Memory...
[ICIP 2022] ACT-NET: Asymmetric Co-Teacher Network for Semi-Supervised Memory...[ICIP 2022] ACT-NET: Asymmetric Co-Teacher Network for Semi-Supervised Memory...
[ICIP 2022] ACT-NET: Asymmetric Co-Teacher Network for Semi-Supervised Memory...
 
[ICIP 2022] MMGL: Multi-Scale Multi-View Global-Local Contrastive learning fo...
[ICIP 2022] MMGL: Multi-Scale Multi-View Global-Local Contrastive learning fo...[ICIP 2022] MMGL: Multi-Scale Multi-View Global-Local Contrastive learning fo...
[ICIP 2022] MMGL: Multi-Scale Multi-View Global-Local Contrastive learning fo...
 
[ICIP 2022 - Poster] MMGL: Multi-Scale Multi-View Global-Local Contrastive le...
[ICIP 2022 - Poster] MMGL: Multi-Scale Multi-View Global-Local Contrastive le...[ICIP 2022 - Poster] MMGL: Multi-Scale Multi-View Global-Local Contrastive le...
[ICIP 2022 - Poster] MMGL: Multi-Scale Multi-View Global-Local Contrastive le...
 
[MICCAI 2022] Meta-hallucinator: Towards Few-Shot Cross-Modality Cardiac Imag...
[MICCAI 2022] Meta-hallucinator: Towards Few-Shot Cross-Modality Cardiac Imag...[MICCAI 2022] Meta-hallucinator: Towards Few-Shot Cross-Modality Cardiac Imag...
[MICCAI 2022] Meta-hallucinator: Towards Few-Shot Cross-Modality Cardiac Imag...
 
[EMBC 2022] Self-supervised Assisted Active Learning for Skin Lesion Segmenta...
[EMBC 2022] Self-supervised Assisted Active Learning for Skin Lesion Segmenta...[EMBC 2022] Self-supervised Assisted Active Learning for Skin Lesion Segmenta...
[EMBC 2022] Self-supervised Assisted Active Learning for Skin Lesion Segmenta...
 
[ICME 2022] Adaptive Mean-Residue Loss for Robust Facial Age Estimation
[ICME 2022] Adaptive Mean-Residue Loss for Robust Facial Age Estimation[ICME 2022] Adaptive Mean-Residue Loss for Robust Facial Age Estimation
[ICME 2022] Adaptive Mean-Residue Loss for Robust Facial Age Estimation
 
[MICCAI 2021] MT-UDA: Towards unsupervised cross-modality medical image segme...
[MICCAI 2021] MT-UDA: Towards unsupervised cross-modality medical image segme...[MICCAI 2021] MT-UDA: Towards unsupervised cross-modality medical image segme...
[MICCAI 2021] MT-UDA: Towards unsupervised cross-modality medical image segme...
 
[EMBC 2021] Multi Slice Dense Sparse Learning for Efficient Liver and Tumor S...
[EMBC 2021] Multi Slice Dense Sparse Learning for Efficient Liver and Tumor S...[EMBC 2021] Multi Slice Dense Sparse Learning for Efficient Liver and Tumor S...
[EMBC 2021] Multi Slice Dense Sparse Learning for Efficient Liver and Tumor S...
 
[ICIP 2020] SEA-Net: Squeeze-and-Excitation Attention Net for Diabetic Retino...
[ICIP 2020] SEA-Net: Squeeze-and-Excitation Attention Net for Diabetic Retino...[ICIP 2020] SEA-Net: Squeeze-and-Excitation Attention Net for Diabetic Retino...
[ICIP 2020] SEA-Net: Squeeze-and-Excitation Attention Net for Diabetic Retino...
 
[MICCAI 2021 - Poster] MT-UDA: Towards unsupervised cross-modality medical im...
[MICCAI 2021 - Poster] MT-UDA: Towards unsupervised cross-modality medical im...[MICCAI 2021 - Poster] MT-UDA: Towards unsupervised cross-modality medical im...
[MICCAI 2021 - Poster] MT-UDA: Towards unsupervised cross-modality medical im...
 

Recently uploaded

Harnessing Passkeys in the Battle Against AI-Powered Cyber Threats.pptx
Harnessing Passkeys in the Battle Against AI-Powered Cyber Threats.pptxHarnessing Passkeys in the Battle Against AI-Powered Cyber Threats.pptx
Harnessing Passkeys in the Battle Against AI-Powered Cyber Threats.pptx
FIDO Alliance
 
Tales from a Passkey Provider Progress from Awareness to Implementation.pptx
Tales from a Passkey Provider  Progress from Awareness to Implementation.pptxTales from a Passkey Provider  Progress from Awareness to Implementation.pptx
Tales from a Passkey Provider Progress from Awareness to Implementation.pptx
FIDO Alliance
 
Cloud Frontiers: A Deep Dive into Serverless Spatial Data and FME
Cloud Frontiers:  A Deep Dive into Serverless Spatial Data and FMECloud Frontiers:  A Deep Dive into Serverless Spatial Data and FME
Cloud Frontiers: A Deep Dive into Serverless Spatial Data and FME
Safe Software
 
TrustArc Webinar - Unified Trust Center for Privacy, Security, Compliance, an...
TrustArc Webinar - Unified Trust Center for Privacy, Security, Compliance, an...TrustArc Webinar - Unified Trust Center for Privacy, Security, Compliance, an...
TrustArc Webinar - Unified Trust Center for Privacy, Security, Compliance, an...
TrustArc
 

Recently uploaded (20)

Stronger Together: Developing an Organizational Strategy for Accessible Desig...
Stronger Together: Developing an Organizational Strategy for Accessible Desig...Stronger Together: Developing an Organizational Strategy for Accessible Desig...
Stronger Together: Developing an Organizational Strategy for Accessible Desig...
 
Six Myths about Ontologies: The Basics of Formal Ontology
Six Myths about Ontologies: The Basics of Formal OntologySix Myths about Ontologies: The Basics of Formal Ontology
Six Myths about Ontologies: The Basics of Formal Ontology
 
Design and Development of a Provenance Capture Platform for Data Science
Design and Development of a Provenance Capture Platform for Data ScienceDesign and Development of a Provenance Capture Platform for Data Science
Design and Development of a Provenance Capture Platform for Data Science
 
Design Guidelines for Passkeys 2024.pptx
Design Guidelines for Passkeys 2024.pptxDesign Guidelines for Passkeys 2024.pptx
Design Guidelines for Passkeys 2024.pptx
 
Harnessing Passkeys in the Battle Against AI-Powered Cyber Threats.pptx
Harnessing Passkeys in the Battle Against AI-Powered Cyber Threats.pptxHarnessing Passkeys in the Battle Against AI-Powered Cyber Threats.pptx
Harnessing Passkeys in the Battle Against AI-Powered Cyber Threats.pptx
 
JavaScript Usage Statistics 2024 - The Ultimate Guide
JavaScript Usage Statistics 2024 - The Ultimate GuideJavaScript Usage Statistics 2024 - The Ultimate Guide
JavaScript Usage Statistics 2024 - The Ultimate Guide
 
Tales from a Passkey Provider Progress from Awareness to Implementation.pptx
Tales from a Passkey Provider  Progress from Awareness to Implementation.pptxTales from a Passkey Provider  Progress from Awareness to Implementation.pptx
Tales from a Passkey Provider Progress from Awareness to Implementation.pptx
 
Modernizing Legacy Systems Using Ballerina
Modernizing Legacy Systems Using BallerinaModernizing Legacy Systems Using Ballerina
Modernizing Legacy Systems Using Ballerina
 
TEST BANK For Principles of Anatomy and Physiology, 16th Edition by Gerard J....
TEST BANK For Principles of Anatomy and Physiology, 16th Edition by Gerard J....TEST BANK For Principles of Anatomy and Physiology, 16th Edition by Gerard J....
TEST BANK For Principles of Anatomy and Physiology, 16th Edition by Gerard J....
 
The Ultimate Prompt Engineering Guide for Generative AI: Get the Most Out of ...
The Ultimate Prompt Engineering Guide for Generative AI: Get the Most Out of ...The Ultimate Prompt Engineering Guide for Generative AI: Get the Most Out of ...
The Ultimate Prompt Engineering Guide for Generative AI: Get the Most Out of ...
 
Cloud Frontiers: A Deep Dive into Serverless Spatial Data and FME
Cloud Frontiers:  A Deep Dive into Serverless Spatial Data and FMECloud Frontiers:  A Deep Dive into Serverless Spatial Data and FME
Cloud Frontiers: A Deep Dive into Serverless Spatial Data and FME
 
Working together SRE & Platform Engineering
Working together SRE & Platform EngineeringWorking together SRE & Platform Engineering
Working together SRE & Platform Engineering
 
Overview of Hyperledger Foundation
Overview of Hyperledger FoundationOverview of Hyperledger Foundation
Overview of Hyperledger Foundation
 
Continuing Bonds Through AI: A Hermeneutic Reflection on Thanabots
Continuing Bonds Through AI: A Hermeneutic Reflection on ThanabotsContinuing Bonds Through AI: A Hermeneutic Reflection on Thanabots
Continuing Bonds Through AI: A Hermeneutic Reflection on Thanabots
 
ChatGPT and Beyond - Elevating DevOps Productivity
ChatGPT and Beyond - Elevating DevOps ProductivityChatGPT and Beyond - Elevating DevOps Productivity
ChatGPT and Beyond - Elevating DevOps Productivity
 
Intro to Passkeys and the State of Passwordless.pptx
Intro to Passkeys and the State of Passwordless.pptxIntro to Passkeys and the State of Passwordless.pptx
Intro to Passkeys and the State of Passwordless.pptx
 
WSO2 Micro Integrator for Enterprise Integration in a Decentralized, Microser...
WSO2 Micro Integrator for Enterprise Integration in a Decentralized, Microser...WSO2 Micro Integrator for Enterprise Integration in a Decentralized, Microser...
WSO2 Micro Integrator for Enterprise Integration in a Decentralized, Microser...
 
TrustArc Webinar - Unified Trust Center for Privacy, Security, Compliance, an...
TrustArc Webinar - Unified Trust Center for Privacy, Security, Compliance, an...TrustArc Webinar - Unified Trust Center for Privacy, Security, Compliance, an...
TrustArc Webinar - Unified Trust Center for Privacy, Security, Compliance, an...
 
WSO2's API Vision: Unifying Control, Empowering Developers
WSO2's API Vision: Unifying Control, Empowering DevelopersWSO2's API Vision: Unifying Control, Empowering Developers
WSO2's API Vision: Unifying Control, Empowering Developers
 
2024 May Patch Tuesday
2024 May Patch Tuesday2024 May Patch Tuesday
2024 May Patch Tuesday
 

[EMBC 2021] Hierarchical Consistency Regularized Mean Teacher for Semi-supervised 3D Left Atrium Segmentation

  • 1. Hierarchical Consistency Regularized Mean Teacher for Semi-supervised 3D Left Atrium Segmentation Shumeng Li*1,2, Ziyuan Zhao*1,2 , Kaixin Xu2, Zeng Zeng2, Cuntai Guan1 Presenter: Zhao Ziyuan 1 Nanyang Technological University 2 Institute for Infocomm Research, A*STAR
  • 2. Background Natural Images Medical Images Need different pre-processing methods Not many classes To segment certain specific structures or tissues ● Image segmentation -> assign a label to every pixel
  • 3. Problem Statement Challenge for medical images: ● Requires manual identification by experts ● Annotations are difficult to obtain Semi-supervised learning Full-supervised learning Data Data Labeled data Labeled data Unlabeled data Improve the segmentation performance with the unlabeled data
  • 4. Proposed Method - Overview 16 128 64 32 16 32 64 128 256 Lc0 Lc1 Lc2 Lc3 Lunsu p Ls0 Ls1 Ls2 Ls3 Lsu p Noise 𝞷 Noise 𝞷’ Teacher Network concatenate downsample conv upsample deconv upsample, conv 1×1, softmax Ground Truth Upsample × 2 Upsample × 4 Upsample × 8 Prediction Input 16 128 64 32 16 32 64 128 256 Student Network Upsample × 2 Upsample × 4 Upsample × 8 EMA (a) Multi-scale deeply supervised network (b) Hierarchical consistency regularized mean teacher
  • 5. Proposed Method (a) - Multi-scale deeply supervised network Ls0 Ls1 Ls2 Ls3 Lsu p concatenate downsample conv upsample deconv upsample, conv 1×1, softmax Ground Truth Prediction 16 128 64 32 16 32 64 128 256 V-Net Upsample × 2 Upsample × 4 Upsample × 8 ● Hierarchical supervised (HS) component ○ add the auxiliary layers after each block in the decoding stage of V-Net ■ an upsampling layer, a 1x1x1 convolution, and a softmax layer ○ multiple prediction results form multi-scale deep supervision ● Supervised loss ○ dice loss + cross-entropy loss
  • 6. Lsu p Lunsu p Proposed Method (b) - Basic mean teacher (MT) 16 128 64 32 16 32 64 128 256 Noise 𝞷 Noise 𝞷’ Teacher Network Ground Truth Prediction Input 16 128 64 32 16 32 64 128 256 Student Network EMA concatenate downsample conv upsample deconv upsample, conv 1×1, softmax ● Semi-supervised learning ○ objective function: supervised loss Lsup + unsupervised consistency loss Lunsup the teacher network weights are updated with the exponential moving average (EMA) of the student network weights Labeled & Unlabeled Labeled Tarvainen, A., & Valpola, H. (2017). Mean teachers are better role models: Weight-averaged consistency targets improve semi- supervised deep learning results. arXiv preprint arXiv:1703.01780.
  • 7. Proposed Method (b) - Hierarchical consistency regularized MT 16 128 64 32 16 32 64 128 256 Lc0 Lc1 Lc2 Lc3 Lunsu p Noise 𝞷 Noise 𝞷’ Teacher Network concatenate downsample conv upsample deconv upsample, conv 1×1, softmax Upsample × 2 Upsample × 4 Upsample × 8 Prediction Input 16 128 64 32 16 32 64 128 256 Student Network Upsample × 2 Upsample × 4 Upsample × 8 EMA ● Hierarchical unsupervised (HU) component ○ add the auxiliary layers to both student and teacher networks ○ further encourage the prediction consistent in the hidden feature space
  • 8. Proposed Method (b) - Hierarchical consistency regularized MT 16 128 64 32 16 32 64 128 256 Lc0 Lc1 Lc2 Lc3 Lunsu p Noise 𝞷 Noise 𝞷’ Teacher Network concatenate downsample conv upsample deconv upsample, conv 1×1, softmax Upsample × 2 Upsample × 4 Upsample × 8 Prediction Input 16 128 64 32 16 32 64 128 256 Student Network Upsample × 2 Upsample × 4 Upsample × 8 EMA ● Consistency loss ○ mean squared error (MSE) loss of multi-scale predictions between the student and the teacher networks
  • 9. Proposed Method - Overview 16 128 64 32 16 32 64 128 256 Lc0 Lc1 Lc2 Lc3 Lunsu p Ls0 Ls1 Ls2 Ls3 Lsu p Noise 𝞷 Noise 𝞷’ Teacher Network concatenate downsample conv upsample deconv upsample, conv 1×1, softmax Ground Truth Upsample × 2 Upsample × 4 Upsample × 8 Prediction Input 16 128 64 32 16 32 64 128 256 Student Network Upsample × 2 Upsample × 4 Upsample × 8 EMA Labeled & Unlabeled Labeled
  • 10. Dataset ● Data source: ○ provided by 2018 Atrial Segmentation Challenge ● Data size: ○ training Set – 80 3D gadolinium-enhanced MRIs and left atrium cavity labels ■ 16 samples - labeled dataset ■ 64 samples - unlabeled dataset ○ test Set – 20 3D MRIs and left atrium cavity labels ● Pre-processing: ○ all the scans were cropped centering at the heart region and normalized as zero mean and unit variance
  • 11. Results ● Our approach achieves the best performance ○ over the state-of-the-art semi-supervised methods in dice, jaccard and 95HD ● Our approach effectively utilizes the unlabeled data ○ compared with full annotation, it obtains comparable segmentation results using only 1/5 labeled data Table 1: Segmentation results of different approaches
  • 12. Results UA-MT SASSNet LG-ER-MT DTC Ours GT ● Example of 2D and 3D results of five semi-supervised segmentation methods, where GT denotes corresponding ground truth labels
  • 13. Results ● To evaluate the effectiveness of HS and HU ○ HS = hierarchical supervised component ○ HU = hierarchical unsupervised component ● Introduce HS to V-Net ○ better utilize the labeled data ● Introduce HU + HS to MT ○ both are effective and combining together can achieve better performance Table 2: Effectiveness of core component in the proposed framework
  • 14. Conclusion ● The proposed framework based on the mean teacher architecture ○ it encourages the consistency of predictions between the student and the teacher networks with the same input under different perturbations ● In each iteration, the student network is optimized by multi-scale deep supervision and hierarchical consistency regularization, concurrently ● Experimental results demonstrate the superiority over the other state-of-the-art SSL methods ● The proposed method is simple and versatile ○ can be widely applied to other segmentation tasks