Estimating respiratory effort through diaphragmatic electromyography on simulated airway obstructions

Obstructive sleep apnea (OSA) is characterized by recurrent airflow interruptions during sleep, leading to increased respiratory effort and significant health risks. Currently, the diagnosis is mainly based on nasal flow, oxygen saturation, and respiratory effort, where the gold standard for the latter, is obtained by placing a catheter in the esophagus to estimate the esophageal pressure (Pes). This study explores diaphragm electromyography (dEMG) as a non-obtrusive alternative to estimate respiratory effort. A dataset including Pes and 6-channel dEMG acquired from healthy volunteers using controlled obstructions was used to obtain the relation between Pes and dEMG in a controlled environment. A number of dEMG features (signal envelope, area under the peak - AUC, peak-to-peak amplitude - PTP and mean power frequency - MNF) were used to quantify the respiratory effort and the results were compared with Pes-based estimates. Significant fair to moderately strong correlations (r>0.5) across all degrees of obstruction were identified between Pes and dEMG envelopes. Moreover, AUC, PTP, and MNF showed comparable agreement between dEMG and Pes (p>0.05) as well as an increasing trend when the resistance to the airflow (obstruction) increased. In conclusion, our investigation of the relationship between dEMG and the corresponding respiratory effort in a controlled lab setting suggests the feasibility of using dEMG signal analysis for the assessment of respiratory dynamics.