| dc.description.abstract | With the modern technology, people can still alive by the assistance of biomedical instrument even when they are seriously injured. Now, many types of ventilators have alarm systems in order to send alert signal to health care personnel when patients are in emergency condition. But the traditional ventilators alarm systems often give false alarm because the criterion was often only relying on one parameter. And health care personnel often can not recognize what the exactly conditions is in the beginning. So they must rely on a wealth of clinical experience to make right judgments after going to patients’ ward. According to that, we design an intelligent alarm system that will help health care personnel work more efficiently. In this study, we design a sputum alarm system to send alert signals when patients’ airway were obstructed by sputum. Until now, health care personnel have to go to patient’s ward to check what patient’s condition is after hearing high pressure alarm. And once they hear something like grunt sounds, they will implement sputum suction process. Thus, we can add that way to our alarm system. The grunt sounds will reflect on flow parameter in the ventilator. If there are some grunt sounds from the patient’s airway, flow wave will reveal some vibrations known as saw tooth pattern. In this study, we quantify the vibrations of the flow wave, and let this value as a criterion that our system can give the exact condition when patient’s was during airway sputum obstruction period. When the value we quantified was high, our system will send airway sputum obstruction alarm. In contrast, when the value was low, our system will send airway static obstruction alarm. | en_US |