Ip based patient monitoring and evaluation system

Real time measuring, monitoring and recording of basic physiological parameters which reflect peoples health situation has an indispensable significance about diagnosing and monitoring of health process. At the present day, devices used for this purpose is being designed more flexible and can be transferred to process of production in paralel with rapid development of embedded system solutions which computing power is high and power consumption is low.Patient monitoring systems known as bedside patient monitor is one of the devices, in this context. This systems can be used to monitor just one patient they can also be used as a module integrated to a central monitoring system in a network configuration. According to this, connecting patient monitor to the central monitoring system over an IP-based communication link has a lot of advantages for purposes of patients and medical personnel. For instance, patient can be monitored in real time from a remote place in hospital or out of hospital by a doctor and/or a nurse and long term assessments can be made about progress of the illness by recording past information on a database.In this thesis, such an IP based patient monitoring system have been designed together with bedside patient and remote monitor units. Hardware substructure of the system, has been composed around a personal computer by using existing embedded system solutions concerning components that forms patient side. Needed interfaces and other software components have been designed on PC. Thus contribution has been given to a potential formation of designation and production of such devices which are in use widely in health sector of our country.Moreover, originality of this thesis, by constituting central process unit which forms unit of patient bedside monitoring of the system has been made an flexible sign process module where signs recorded on PC can be evaluated according to different algorithms. On this module, by processing signs recorded in the system such as respiratory (RESP, SpO2) and electrocardiograph (ECG) in real time, different algorithms have been suggested and tested for diagnosis of apnea. Because of difficulties on finding adequate healthy and sick subjects for evaluations on this purpose, signs taken from an existing database was used.Assessments made on respiratory scores recorded directly from `abdomen? rate of success which reaches %94 has been gathered in real time detection of apnea in minute scale. As a result of evaluation of different respiratory scores subjects apneic and non-apneic have been determined with accuracies betweeen %93,33 - %96,66, on subject based assesments. Such a design, at the same time, gives support to studies developing a product which can be used as an alternative to `polysomnography? which is currently in widely use for diagnosis of apnea.