Artificial Intelligence signal processing for enhancing an intelligent sensor

Abstract

This report has summarized and recorded on what have been done on the whole process of this final year project. To overcome and understand the scope of the project, this report will be divided into five main parts. First part will be mainly focus on the scope of the project and clearly underline the objectives which is to learn and better appreciate the working principle of photoacoustic to measure blood glucose in a non-invasive method. Second part we will move our attention to signal processing, we will learn the basic analogy of signals such as, type of signals, what is signals use for and the importance of signal. Slowly we learn three methods of AI signal processing which are Principle Component Analysis, Partial Least Square Regression and Multi Linear Regression. Come to the end of this part we will create a simple PCA coding using MatLab. Third part we will look and explore the photoacoustic world that is currently being under-study by researches all over the world to implement in the bio-medical field as a non-invasive method in doing things. Two different types of photoacoustic model will be touch in this report which are bipolar ”Nshape” pulse from an optical absorber and PA damping oscillation effect. Last two parts will be touching on the specification of the current prototype that is in the lab and also results of lab experiment. From the lab experiment we will analyse the data by using linear regression and PCA methods of calibration to prove that photoacoustic are still statically and clinically acceptable in this bio-medical field.