MEDICAL CONTROL SYSTEMS

Author: RAYGOZA PANDURO JUAN JOSE.
Year: 2004.
University: AUTÓNOMA DE MADRID [www.uam.es].
Place of defense: DEPARTAMENTO DE INFORMATICA.
Place of preparation: ESCUELA POLITECNICA SUPERIOR.

Author: RODRÍGUEZ ARRIETA FERNANDA JIMENA.
Year: 2005.
University: PAÍS VASCO [www.ehu.es].
Place of defense: FACULTAD DE INFORMÁTICA DE SAN SEBASTIÁN.

Summary: It has done the analysis, design, development and implementation of a set of tools aimed at telemonitorización ubiquitous, intelligent and continuous sick heart, react and prevent serious cardiac arrhythmias. It presents a formal framework for analyzing signals through a set of techniques drawn from the area of automatic learning with the goal of obtaining the best and most complete classifier beats and rhythms. It also has implemented a prototype molec: a system telemonitorización ubiquitous, intelligent, efficient and continuous in the environment of mobile computing, that takes on a continuous signals heart beats and identifies its components in real time, classifies beats and rhythms, identifies and sends alarms, stores and manages data in cardiology locally and remotely. Finally provides a means of access to information classified and managed remotely on a mobile device using web services technology.

Author: ALESANCO IGLESIAS ALVARO.
Year: 2006.
University: ZARAGOZA [www.unizar.es].
Place of defense: CENTRO POLITÉCNICO SUPERIOR.
Place of preparation: CENTRO POLITÉCNICO SUPERIOR.

Summary: This thesis presents the design, implementation and evaluation of a complete system of telecardiogía time. Right. The first part is treated encoding in real time with wavelet transform and the algorithm SPIHT of ECGs that will lead to losses in the quality of ECG scrambled. Since a loss of quality can lead to excessive signal not clinically useful, the distortion introduced in the encoding process must be evaluated from a clinical point of view. One of the most important contributions of this part is that compression is done automatically depending on the noise that contains the block or beat that is scrambled. Thanks to this contribution, ensures quality clinical reconstruction at the same time eliminates the excessive noise of the ECG signal. To validate the clinical quality of the compressed signal, have designed two clinical test. Three cardiologists have participated in the clinical evaluation of the signals compressed using the approximation automatic threshold of noise. The results of the clinical indices for signals evaluated shed excellent results: the quality of the signals compressed clinic is described as very good, the highest rating possible. The second part deals with the transmission of ECGs coded on communications networks and the effects of transmission errors on the monitoring of ECG at the reception. To that end, it has proposed the creation and use another application layer protocol: Reliable ECG Transmission Protocol (RETP). This protocol is responsible for conducting retransmissions of the lost information on the network. The study in wide-area networks without wires has shown how increasing the Bit Error Rate (BER) and the Doppler frequency (f _d $ $) channel have a negative effect on ECG monitoring in real time, given that increase in the number of stops per minute and the average length of the same experience that the process of monitoring. Moreover, it also increases the delay ECG monitoring, which is also very harmful to watch. These effects have been discussed for two different sizes buffer watch: 2 sy 3 s, watching as a buffer greater mitigates the harmful effects of the loss and retransmission of packets thing to introduce a delay Increased monitoring. The study in wired networks WAN show how the increase in the average length of situations congestion on the network as well as the increase in the total percentage of time that the channel is in congestion exact a heavy toll on the quality of monitoring, making both the number of stops, their average length of the delay and increase monitoring. Finally, in order to extrapolate the results of number of stops, the same average length of delay and watch to a clinic, a study has been made with cardiologists in which it has assessed the monitoring of ECGs received different percentage of time stopped monitoring (parameter is derived from the number of stops and the average length), concluding that if the monitoring is clinically acceptable, this percentage should be lower than 15%. Moreover, another study showed that the cardiologists retardation monitoring should be less than 3.3 s to be able to see the signal with confidence. Taking into account all these results, it is easy to extrapolate them to the results obtained in the various networks studied and delineate the conditions of the channel in which the transmission will be feasible from a clinical point of view.