Patient stratification and clinical decision support on mechanical ventilation settings from sonar responses through an endotracheal tube (ETT)

A respiration monitoring device comprising an electronic controller configured to:

• receive an audio signal that is acoustically coupled with an airway of a patient receiving mechanical ventilation therapy from a mechanical ventilator;
• map resonant frequencies extracted from the audio signal to one or more lung disease or injury condition categories;
• and
• at least one of:
• display the mapped one or more lung disease or injury condition categories on a display device;
• and
• determine a recommended adjustment to one or more parameters of the mechanical ventilation therapy delivered to the patient based at least on the mapped lung disease or injury condition categories and displaying the recommended adjustment on the display device;
• wherein the electronic controller is configured to map the resonant frequencies to the one or more lung disease or injury condition categories by inputting the resonant frequencies to a resonant frequencies-to-lung disease or injury condition mapping machine-learning (ML) model, wherein the ML model is trained by:
• obtaining training images of a plurality of historical patients wherein the training images are labeled with lung disease or injury categories of the historical patients;
• simulating audio signals for the training images to generate simulated audio signals labeled with lung disease or injury categories;
• and
• training the resonant frequencies-to-lung disease or injury condition mapping ML model using the simulated audio signals labeled with lung disease or injury categories.