Ultrasound‐Based Local Lung Motion Assessment Using Synthetic Lateral Phase

TL;DR

This paper introduces a new ultrasound-based method to track lung motion, which could improve the evaluation of lung diseases.

Contribution

A novel synthetic lateral phase-based algorithm for tracking lung motion is developed and validated.

Findings

The synthetic lateral phase-based algorithm achieved 3% accuracy in simulated lung motion.

Lung displacement was greatest in lower and lateral zones in both healthy and COVID-19 subjects.

Regular breathing showed greater displacement than rapid shallow or breath-hold patterns.

Abstract

Ultrasound lung surface motion measurement is valuable for the evaluation of a variety of diseases. Speckle tracking or Doppler‐based techniques are limited by the loss of visualization as a tracked point moves under ribs or is dependent.

We developed a synthetic lateral phase‐based algorithm for tracking lung motion to overcome these limitations. To validate the technique, we generated simulated lung motion images. We also obtained lung ultrasound cines from a healthy volunteer and a mechanically ventilated COVID‐19 patient. In the healthy volunteer, the respiratory pattern varied between breath‐hold, regular, and rapid shallow breathing.

The measured displacement was within 3% of the ground truth for simulated cines. In both the healthy volunteer and COVID‐19 patients, measured displacement was greatest in the lower and lateral zones of the lung when the ipsilateral side was compared. In the healthy volunteer, when the respiratory pattern was varied, measured displacement was greater in regular breathing compared to rapid shallow breathing and compared to breath‐hold patterns in both the upper and lower lung zones.

Estimation of lung surface displacement using a synthetic lateral phase‐based approach is feasible. Future human studies should validate this approach against a direct measurement of lung surface movement.

Representative workflow images from lung motion tracking software. (A) Screen capture of user marking lung surface. (B) Screen capture of displacement heatmap and selection of region of interest for motion tracking. (C) Video of motion tracking. (D) Screen capture of breath marking.

Figures