Abstract: A noninvasive of detecting patient-ventilator asynchrony that is easily adaptable to existing ventilator monitoring systems and provides timely and actionable information on the degree of patient asynchrony both during invasive and non-invasive ventilation. Capture, analysis or display of, frequency spectra and the use of a measure of spectral organization, such as H1/DC, allows for both manual and automatic adjustment of a ventilators to prevent or correct patient-ventilator asynchrony via interventions. Embodiments use artificial intelligence or machine learning to predict interventions predicted to result in positive outcomes, based on analysis of a large number of epochs, captured by an electronic monitor of a mechanical ventilator, where the monitor continuously monitors, captures and transfers, epochs of data for aggregated machine learning analysis, of such epochs associated with positive outcomes.