Skip to main content
Fig. 2 | Respiratory Research

Fig. 2

From: Deep learning for spirometry quality assurance with spirometric indices and curves

Fig. 2

Flowchart of the algorithm training and testing data acquisition, selection, and division. 16,502 spirometry files were retrieved from the First Affiliated Hospital of Guangzhou Medical University. After exclusion of files with no curves or curves that could not be resolved, 15,693 files remained, files were randomly divided into training and internal testing sets. Additional 219 spirometry files retrieved from the multicenter (three hospitals) of the National Clinical Research Center for Respiratory Disease were used for external testing. FEV1  forced expiratory volume in 1 s, FVC forced vital capacity

Back to article page