DUCPS: Deep Unfolding the Cauchy Proximal Splitting Algorithm for B-Lines Quantification in Lung Ultrasound Images

TL;DR

This paper introduces DUCPS, an unsupervised deep unfolding network that enhances B-line detection in lung ultrasound images by combining traditional algorithms with deep learning, achieving faster and more accurate results without extensive data labeling.

Contribution

The paper presents a novel unsupervised deep unfolding framework that improves B-line detection in lung ultrasound images by integrating traditional model-based methods with deep learning.

Findings

Achieved state-of-the-art recall and F2 scores of 0.70 and 0.64.

Significantly improved computational efficiency over traditional and existing deep learning methods.

Enabled rapid convergence and better generalization through learnable parameters.

Abstract

The identification of artefacts, particularly B-lines, in lung ultrasound (LUS), is crucial for assisting clinical diagnosis, prompting the development of innovative methodologies. While the Cauchy proximal splitting (CPS) algorithm has demonstrated effective performance in B-line detection, the process is slow and has limited generalization. This paper addresses these issues with a novel unsupervised deep unfolding network structure (DUCPS). The framework utilizes deep unfolding procedures to merge traditional model-based techniques with deep learning approaches. By unfolding the CPS algorithm into a deep network, DUCPS enables the parameters in the optimization algorithm to be learnable, thus enhancing generalization performance and facilitating rapid convergence. We conducted entirely unsupervised training using the Neighbor2Neighbor (N2N) and the Structural Similarity Index Measure (SSIM) losses. When combined with an improved line identification method proposed in this paper, state-of-the-art performance is achieved, with the recall and F2 score reaching 0.70 and 0.64, respectively. Notably, DUCPS significantly improves computational efficiency eliminating the need for extensive data labeling, representing a notable advancement over both traditional algorithms and existing deep learning approaches.