Automatic Algorithm for Quality Assessment of the Unsupervised Spirometry Based on Machine Learning Method

Abstract

Home monitoring of the lung function using handheld spirometers have many benefits including early diagnosis, prediction of upcoming exacerbations or treatment optimization. However, the quality of the waveforms obtained from unsupervised spirometry requires appropriate assessment before reliable interpretation, even if they meet all ATS/ERS criteria. In this study, we present an automatic algorithm for quality assessment of the spirometry based on machine learning.