Automatic detection of patients' spontaneous activity during pressure support ventilation

The occurrence of significant patient-ventilator asynchronies in assisted ventilation modes is an impellent problem in clinical practice. Addressing this question, an original software has been developed and is here proposed. This tool implements a new automatic technique to identify the beginning and the end of the patient’s respiratory effort, events that are sometimes missed or detected with significant delay by the ventilator. Its improved skills have been evaluated on a set of signals coming from 6 ICU patients and including 6445 respiratory acts, and proved to outperform the machine in increasing the amount of respiratory acts assisted without significant delay from 22 to 70%. The presented tool is the first step in the development a hardware-software device to be directly interfaced with the ventilator in order to represent a monitoring aid for the clinician and possibly to directly drive the device activity.