In daily life every person needs continuous monitoring of temperature, heart rate, oxygen saturation level, blood pressure parameters and other parameters to have some idea about one's body systems performance and to assist doctors to diagnose one's health status. This information is more necessary for aged and unhealthy people, while it is also necessary for healthy person, who represents the undiagnosed subject.
Usually, the healthy and unhealthy subjects are advised to measure their cardiovascular parameters at home at various times in a day to avoid any bad developments for their health status. Available self-measurement devices give only discrete readings and have not provided accurate information of heart rate, oxygen saturation level, and blood pressure parameters in many situations since most of them do not consider the body's movement or the uncertainty associated with the reading.
Moreover, Blood pressure parameters (BPP): Systolic, Diastolic, and Mean Blood Pressures, have some types of correlation with the heart rate. This relationship is nonlinear and has many levels of uncertainty. The Type-2 Fuzzy system has a capacity to deal with nonlinear and uncertainty systems. The estimate of Blood pressure parameters based on heart rate can be categorized under such systems that fuzzy system can deal with.
This thesis presents a novel algorithm to measure photo-plethysmography signal, heart rate and the oxygen saturation level and also to estimate BPPs for healthy and unhealthy subjects based on a prototype transducer, particle swarm optimization and type-2 Fuzzy System.
The measured values of heart rate, oxygen saturation level, systolic, diastolic and mean blood pressures by utilizing the novel algorithn1 are compared with the clinical readings of heart rate, oxygen saturation level, systolic, diastolic and mean blood pressure.
Very encouraging results have been achieved for estimating heart rate, oxygen saturation level, systolic, diastolic and mean blood pressures and the accuracy of estimated results for that parameters for healthy subjects, by our designed algorithm, are 99.53%,98.91 %,97.76%,91.81 % and 96.43%, respectively.
Add to that, the accuracy of estimated results systolic, diastolic and mean arterial blood pressure for unhealthy subjects are 94.51 %,91.48% and 94.79%, respectively.
On the contrary, the mean arterial blood pressure is estimated based on same heart rate and existing algorithm. This algorithm can only estimate mean arterial blood pressure. The accuracy of estimated mean arterial blood pressure equals 53.83%.
The proposed model achieves very encouraging results; since all accuracies of the blood pressure parameters for unhealthy and healthy subjects are more than 91.4%. Moreover, the proposed algorithm can be utilized to determine heart rate, oxygen saturation level, systolic, diastolic and mean blood pressures.