A recurrent neural network classifier achieved expert-level performance in detecting ventilatory thresholds from CPET data, with mean absolute errors of 9.5% (r=0.79) for VT1 and 4.2% (r=0.94) for VT2.
Does a recurrent neural network accurately detect ventilatory thresholds from CPET data compared to expert visual assessment?
A recurrent neural network can accurately detect ventilatory thresholds from CPET data, achieving performance comparable to expert exercise physiologists.
Effect estimate: MAE 9.5% (r=0.79) for VT1; MAE 4.2% (r=0.94) for VT2
Abstract First and second ventilatory thresholds (VT 1 and VT 2 ) represent the boundaries of the moderate‐heavy and heavy‐severe exercise intensity. Currently, VTs are primarily detected visually from cardiopulmonary exercise test (CPET) data, beginning with an initial data screening followed by data processing and statistical analysis. Automated VT detection is a challenging task owing to the high signal to noise ratio typical of CPET data. Recurrent neural networks describe a machine learning form of Artificial Intelligence that can be used to uncover complex non‐linear relationships between input and output variables. Here we proposed detection of VTs using a single neural network classifier, trained with a database of 228 laboratory CPET data. We tested the neural network performance against the judgement of 7 couples of board‐certified exercise‐physiologists on 25 CPET tests. The neural network achieved expert‐level performances across the tasks (mean absolute error was 9.5% ( r = 0.79) and 4.2% ( r = 0.94) for VT 1 and VT 2 , respectively). Estimation errors are compatible with the typical error of the current gold standard visual methodology. The neural network demonstrated VT detecting and exercise intensity level classifying at a high competence level. Neural networks could potentially be embedded in CPET hardware/software to extend the reach of exercise physiologists beyond their laboratories.
Zignoli et al. (Mon,) conducted a other in Cardiopulmonary exercise test (CPET) data (n=253). Recurrent neural network classifier vs. Visual detection by board-certified exercise-physiologists was evaluated on Mean absolute error and correlation (r) for first and second ventilatory thresholds (VT1 and VT2) (MAE 9.5% (r=0.79) for VT1; MAE 4.2% (r=0.94) for VT2). A recurrent neural network classifier achieved expert-level performance in detecting ventilatory thresholds from CPET data, with mean absolute errors of 9.5% (r=0.79) for VT1 and 4.2% (r=0.94) for VT2.