Geometric Analyses of the Expiratory Flow–Volume Curve to Identify Expiratory Flow Limitation During Exercise

• Published in: Fluids (Basel) Vol. 10; no. 4; p. 94
• Main Authors: Haverkamp, Hans, Petrics, Gregory, Molgat-Seon, Yannick
• Format: Journal Article
• Language: English
• Published: Basel MDPI AG 01.04.2025
• Subjects: Aerobics; Airway management; Analysis; Artificial neural networks; breathing mechanics; cardiopulmonary exercise testing; Chronic obstructive pulmonary disease; deep learning; Dyspnea; Exercise; exertional dyspnea; Gender differences; Homeostasis; Lungs; Machine learning; Metabolism; Neural networks; Older people; Patient compliance; Physical fitness; Physiological aspects; Physiology; Respiratory system; Tidal flow
• ISSN: 2311-5521; 2311-5521
• DOI: 10.3390/fluids10040094

An important purpose of cardiopulmonary exercise testing (CPET) is to query the mechanisms for unexplained shortness of breath or exaggerated exertional dyspnea. Expiratory flow limitation (EFL) is an important indicator of ventilatory constraint that can negatively influence both dyspnea and exercise capacity. Unfortunately, due to logistical challenges and lack of sufficient clinical training, EFL is rarely measured during CPET. The conventional method for identifying exercise EFL is limited because it requires patient cooperation and it is also dependent on the maximal expiratory flow–volume curve, which underestimates actual maximal expiratory flow during exercise. Simplified methods for identifying EFL that are based on the shape of the exercise tidal flow–volume curve would improve the accessibility of measuring EFL during exercise. The overall aim of this review is to critically review the approaches and methods used to measure EFL in exercising adults. We review the physiology underlying EFL and the conventional methods for determining exercise EFL. We then provide critical analyses of more recent methods for identifying exercise EFL that are based on the geometry of the exercise tidal expiratory flow–volume curve. Finally, we highlight recent work designed to assess exercise EFL using a type of deep machine learning known as a convolutional neural network.