Process and signal processing unit for determining a pneumatic parameter with the use of a lung-mechanical model and of a gradient model

A signal processing unit for an approximate automatic determination by calculation of a pneumatic parameter for spontaneous breathing of a patient, the signal processing unit comprising:

• a processor comprising at least occasional read access to a memory with a stored computer-analyzable lung-mechanical model and a stored computer-analyzable gradient model for the pneumatic parameter, wherein:
• the lung-mechanical model describes at least one relationship between the pneumatic parameter and one or more of a volume flow signal for a flow of breathing air to and/or from the lungs of the patient, and a volume signal for a filling level of the lungs of the patient and a measurable pneumatic pressure signal for pressure in the airways or in an esophagus of the patient and a respiratory signal which is correlated with the spontaneous breathing of the patient;
• the gradient model describes a value of the pneumatic parameter for a scanning time at least as a function of N values of a variable, which is variable over time and is correlated with the pneumatic parameter for N previous scanning times, wherein N is a preset number;
• the processor is configured to generate a respective signal value repeatedly for at least one signal to which the lung-mechanical model refers,
• the processor is configured to receive measured values from at least one of an airway pressure sensor and an esophageal pressure sensor and a volume flow sensor and a volume sensor to generate from the received measured values the volume flow signal and the volume signal and the pressure signal and measured values from a sensor array which measures a variable that is correlated with the spontaneous breathing of the patient, and a second sensor array on the skin of the patient and from a sensor in the body of the patient when generating the signal values;
• the processor is configured to determine, during an initialization phase, N values of the variable or each variable correlating with the pneumatic parameter and occurring in the gradient model for N different scanning times;
• the processor is further configured to determine in a subsequent use phase a respective value for the pneumatic parameter for at least one scanning time,
• the processor is configured to use for this determination during the subsequent use phase at least:
• generated signal values for the at least one scanning time;
• respective, already determined N values of the variable or each variable of the gradient model, which variable is correlated with the pneumatic parameter, for N previous scanning times,
• the lung-mechanical model;
• and
• the gradient model;
• the processor is configured to provide the pneumatic parameter as output to a ventilator for synchronizing operation of the ventilator with the spontaneous breathing of the patient.