Transfer Learning U-Net Deep Learning for Lung Ultrasound Segmentation

TL;DR

This paper compares transfer learning techniques using U-Net for lung ultrasound rib segmentation, demonstrating that pre-training on natural images and data augmentation improve segmentation accuracy and artifact reduction.

Contribution

It introduces and evaluates two transfer learning approaches for U-Net in lung ultrasound segmentation, highlighting the effectiveness of pre-training and fine-tuning strategies.

Findings

X-Unet achieved more accurate visual results despite lower DICE scores.

Full fine-tuning slightly outperformed partial freezing.

Dataset size and data augmentation significantly impact performance.

Abstract

Transfer learning (TL) for medical image segmentation helps deep learning models achieve more accurate performances when there are scarce medical images. This study focuses on completing segmentation of the ribs from lung ultrasound images and finding the best TL technique with U-Net, a convolutional neural network for precise and fast image segmentation. Two approaches of TL were used, using a pre-trained VGG16 model to build the U-Net (V-Unet) and pre-training U-Net network with grayscale natural salient object dataset (X-Unet). Visual results and dice coefficients (DICE) of the models were compared. X-Unet showed more accurate and artifact-free visual performances on the actual mask prediction, despite its lower DICE than V-Unet. A partial-frozen network fine-tuning (FT) technique was also applied to X-Unet to compare results between different FT strategies, which FT all layers slightly outperformed freezing part of the network. The effect of dataset sizes was also evaluated, showing the importance of the combination between TL and data augmentation.