B-line Detection in Lung Ultrasound Videos: Cartesian vs Polar Representation

TL;DR

This paper introduces an attention-based Convolutional+LSTM model for automatic B-line detection in lung ultrasound videos, comparing Cartesian and polar image representations, and finds polar representation offers competitive performance and better localization.

Contribution

The study proposes a novel deep learning framework utilizing attention mechanisms for B-line detection and evaluates the effectiveness of polar versus Cartesian representations.

Findings

Polar representation achieves comparable B-line classification accuracy to Cartesian.

Attention mechanism improves localization of B-lines.

Polar representation may better capture radial artefacts in LUS images.

Abstract

Lung ultrasound (LUS) imaging is becoming popular in the intensive care units (ICU) for assessing lung abnormalities such as the appearance of B-line artefacts as a result of severe dengue. These artefacts appear in the LUS images and disappear quickly, making their manual detection very challenging. They also extend radially following the propagation of the sound waves. As a result, we hypothesize that a polar representation may be more adequate for automatic image analysis of these images. This paper presents an attention-based Convolutional+LSTM model to automatically detect B-lines in LUS videos, comparing performance when image data is taken in Cartesian and polar representations. Results indicate that the proposed framework with polar representation achieves competitive performance compared to the Cartesian representation for B-line classification and that attention mechanism can provide better localization.