# EDMAE: An Efficient Decoupled Masked Autoencoder for Standard View Identification in Pediatric Echocardiography

## TL;DR

EDMAE is a self-supervised convolutional autoencoder that efficiently recognizes standard pediatric echocardiography views, outperforming existing methods in accuracy and downstream tasks.

## Contribution

Introduces EDMAE, a novel self-supervised autoencoder with a teacher-student structure using convolutional encoders for efficient view recognition in pediatric echocardiography.

## Key findings

- Achieves high accuracy on 27 standard views.
- Outperforms popular supervised and self-supervised methods.
- Effective in downstream cardiac ultrasound segmentation.

## Abstract

This paper introduces the Efficient Decoupled Masked Autoencoder (EDMAE), a novel self-supervised method for recognizing standard views in pediatric echocardiography. EDMAE introduces a new proxy task based on the encoder-decoder structure. The EDMAE encoder is composed of a teacher and a student encoder. The teacher encoder extracts the potential representation of the masked image blocks, while the student encoder extracts the potential representation of the visible image blocks. The loss is calculated between the feature maps output by the two encoders to ensure consistency in the latent representations they extract. EDMAE uses pure convolution operations instead of the ViT structure in the MAE encoder. This improves training efficiency and convergence speed. EDMAE is pre-trained on a large-scale private dataset of pediatric echocardiography using self-supervised learning, and then fine-tuned for standard view recognition. The proposed method achieves high classification accuracy in 27 standard views of pediatric echocardiography. To further verify the effectiveness of the proposed method, the authors perform another downstream task of cardiac ultrasound segmentation on the public dataset CAMUS. The experimental results demonstrate that the proposed method outperforms some popular supervised and recent self-supervised methods, and is more competitive on different downstream tasks.