Contrastive Pretraining for Echocardiography Segmentation with Limited Data

Mohamed Saeed,Rand Muhtaseb,Mohammad Yaqub

DOI: https://doi.org/10.48550/arXiv.2201.07219

2022-07-15

Abstract:Contrastive learning has proven useful in many applications where access to labelled data is limited. The lack of annotated data is particularly problematic in medical image segmentation as it is difficult to have clinical experts manually annotate large volumes of data such as cardiac structures in ultrasound images of the heart. In this paper, We propose a self supervised contrastive learning method to segment the left ventricle from echocardiography when limited annotated images exist. Furthermore, we study the effect of contrastive pretraining on two well-known segmentation networks, UNet and DeepLabV3. Our results show that contrastive pretraining helps improve the performance on left ventricle segmentation, particularly when annotated data is scarce. We show how to achieve comparable results to state-of-the-art fully supervised algorithms when we train our models in a self-supervised fashion followed by fine-tuning on just 5\% of the data. We show that our solution outperforms what is currently published on a large public dataset (EchoNet-Dynamic) achieving a Dice score of 0.9252. We also compare the performance of our solution on another smaller dataset (CAMUS) to demonstrate the generalizability of our proposed solution. The code is available at (<a class="link-external link-https" href="https://github.com/BioMedIA-MBZUAI/contrastive-echo" rel="external noopener nofollow">this https URL</a>).

Image and Video Processing,Computer Vision and Pattern Recognition,Machine Learning

What problem does this paper attempt to address?

This paper attempts to solve the problem of echocardiogram left ventricular segmentation under the condition of limited labeled data. Specifically, the authors propose a self - supervised contrastive learning method to overcome the performance limitations caused by insufficient labeled data in medical image segmentation. This method is especially suitable for ultrasonic images of cardiac structures, because it is very difficult and costly for clinical experts to manually label a large amount of data. By using the contrastive pre - training technique, the paper studies its impact on two known segmentation networks (UNet and DeepLabV3), and shows how to achieve results comparable to those of fully - supervised algorithms with only a small amount of labeled data. The main contribution of the paper lies in proving that contrastive pre - training can significantly improve the performance of left ventricular segmentation, especially in the case of scarce labeled data.

Contrastive Pretraining for Echocardiography Segmentation with Limited Data

An improved contrastive learning network for semi-supervised multi-structure segmentation in echocardiography

Learning to segment with limited annotations: Self-supervised pretraining with regression and contrastive loss in MRI

Contrastive learning of global and local features for medical image segmentation with limited annotations

Efficient deep learning-based automated diagnosis from echocardiography with contrastive self-supervised learning

Self-Supervised Pre-Training Based on Contrastive Complementary Masking for Semi-Supervised Cardiac Image Segmentation

Multi-level Asymmetric Contrastive Learning for Medical Image Segmentation Pre-training

Contrastive Learning for View Classification of Echocardiograms

Self-supervised contrastive learning of echocardiogram videos enables label-efficient cardiac disease diagnosis

Multi-level Asymmetric Contrastive Learning for Volumetric Medical Image Segmentation Pre-training

Contrastive and Selective Hidden Embeddings for Medical Image Segmentation

GLRP: Global and Local Contrastive Learning Based on Relative Position for Medical Image Segmentation on Cardiac MRI

U-shape-based network for left ventricular segmentation in echocardiograms with contrastive pretraining

Keypoint-Augmented Self-Supervised Learning for Medical Image Segmentation with Limited Annotation

An efficient annotated data generation method for echocardiographic image segmentation

Self-Ensembling Contrastive Learning for Semi-Supervised Medical Image Segmentation

Self-Supervised Learning from Unlabeled Fundus Photographs Improves Segmentation of the Retina

Self-supervised Contrastive Learning with Random Walks for Medical Image Segmentation with limited Annotations

Multi-organ Self-supervised Contrastive Learning for Breast Lesion Segmentation

Automatic Cardiac MRI Segmentation Using Deep Learning

Localized Region Contrast for Enhancing Self-Supervised Learning in Medical Image Segmentation