Respiratory trigger signal generation by means of a stretchable sensor array

Abstract

Listen

Translate article

Respiratory monitoring is a clinical method which helps to examine the medical condition of patients. Patients diagnosed with types of respiratory distress are often supported through artificial respiration. To be able to adapt and synchronize airway pressures and flows to the patient’s own breathing for improved respiration efficiency, intelligent sensors are needed to detect the beginning and ending of the breathing cycle. An ultra-thin and stretchable 6 × 6 sensor array with skin-like properties is presented that is used to generate a trigger signal which is suitable to control and synchronize artificial respiration with the patient's own breathing. Stretchability of the sensor array is achieved by fs-laser structuring of the thin polyimide sensor substrate resulting in small sensor islands connected via slender meandering electrical leads. The resulting stretchable sensor grid is embedded in layers of PDMS whereby a skin-friendly sensor patch is created. To simulate respiration an externally ventilated dummy is used. The principle of trigger signal generation from multiple sensor signals is based on a self-developed algorithm that first evaluates the signal quality of each sensor based on adjustable parameters. Only the sensors selected as suitable are then used to calculate an averaged scaled signal, which is taken for trigger point detection. The best results were typically obtained when quality factures are set to a level where about half of the sensors are contributing to the trigger detection, leading to a trigger delay of about 80 ms relative to the pressure reference signal. It could also be shown that the algorithm can resume the trigger point detection within 2–3 s, after manually applying disturbances which could similarly occur in the clinical environment. The results show that the skin-friendly sensor patch provides suitable trigger signals for artificial respiration which are robust against drop out of single sensors, non-ideal sensor patch positioning on the thorax and mechanical irritations.