Rediseño del Prototipo Cardio One V1.0 y Desarrollo de Algoritmos Confiables que Permiten la Identificación de Bio-marcadores a partir de las Señales ECG y PPG

Abstract

Mortality due to cardiovascular system diseases represents a big number of deaths worldwide each year, in Colombia the hearth diseases have the highest annual percentage of deaths. If this fact is attached to low implementation of prevention and early supervision systems, a great chance of innovation is observed. The current project looks for an alternative of ambulatory and no invasive strategy for the cardiovascular system, a wearable device able to monitor hearth variables using the Electrocardiogram signal (ECG) and the blood volume variation by the photoplethysmography technic that provides biomarkers like the cardiac frequency, its variability, and the oxygen percentage in blood with the objective to bring the users a tool for tracing hearth variables in daily activities and sport fields. First, the previous project result was validated, the Cardio One V1.0, to ensure that its instrumentation and operation for the redesign and miniaturization fulfills certain minimum standards of precision and exactitude in the context of the biological variables that is proposed to measure. Later, interviews were made to potential users with the objective to extract the functional criteria for the design and building of the future prototype Cardio One V2.0. The version 1.0 hardware of the device is composed by two plates, first one contains the MAX86150 sensor with its instrumentation and the second one brings the sensor interaction interface, electrode connector, source modules, communication modules, main SoC and other sensors. In the redesign stage it was determined that the device will be conformed in same way, by two PCB plates, but adopting modularity concept, that brings the improvement chance. It was appeal to two- and four-layers PCB, separating the source, communication and main SoC modules from the sensor to optimize space and reduce the dimensions of the device with the objective to ease the coupling to the human body based on concepts of ergonomics and detail design. In the software developing, it was researched which algorithms and methodologies allows the satisfactory noise elimination of the signals that are handled. It was determined that the Chebyshev Two and Butterworth filters with the type three Wavelet transform and medium value elimination, results in a low noise and defined signal. From the obtained signal in the filtering stage biomarkers are extracted. This stage eases the extraction of useful information due to need of complex technics and algorithms that represents long computing times are deleted. With the device Cardio One V2.0 and making good use of the developed algorithms trustworthy measurement of the electrocardiogram is reached, the frequency and hearth variability and oxygen percentage bringing the user the ambulatory tracing of its cardiovascular health chance.