Development and application of multi dynamic analysis in physiological signals

• Main Authors: Yu-Cheng Lin, 林祐正
• Other Authors: Terry B.J. Kuo
• Format: Others
• Language: zh-TW
• Published: 2016
• Online Access: http://ndltd.ncl.edu.tw/handle/43026042051777141367

博士 === 國立陽明大學 === 腦科學研究所 === 104 === With the advance of technology and medicine, wearable physiological sensing device products, cloud healthcare system and internet of things medical wisdom industry is rapidly growing, a variety of physiological signals of the body making easier to measure and store. Because the physiology is ultimate complex system, physiological signal of complex fluctuation not only influence sympathetic and parasympathetic nervous of autonomic nerve system, endocrine, respiratory system and unpredictability factors, but also influence external season, environment, temperature, mood and changes with time on different spatial and temporal scales. However, these phenomena would comprise a complex physiologic signals with multiple dimensions and nonlinear and non-stationary. In fact, the traditional methods for detection of complex dynamic signal still have many limitations, and even difficult to have analytical ability for variation characteristics of complex signals from microscopic to macroscopic level on different time and window length simultaneously. Although the invasive biochemical indicators of the health care system or major medical equipment can be used to examine and evaluate the disease severity and prognosis, there are still have many limitations and inconvenient, and the great amount of equipment cost and professional technical personnel are need to be expended and invited that it difficult to extend of home or personal, and these indicators not have capacity of daily detection. The Hilbert-Huang Transform (HHT) is a novel signal processing method, which consists of Empirical Mode Decomposition (EMD) and Hilbert Spectral Analysis. EMD, the key step of HHT, can adaptively decompose the signal into number of different intrinsic mode functions (IMFs) operated in different time-scales and a non-oscillation trend. The study aims to apply HHT, cloud computing technology and variable slide window method in several biomedical topics including (1) we combine Xenon uploading system, Hilbert Huang Transform and cloud computing to construct the convenient operation and reliable automatic cloud-computing healthcare system for investigation the intrinsic trend change of blood pressure and heart rate; (2) we combine slide window size method and empirical mode decomposition method of HHT to develop new data analysis technique - the Multi Dynamic Trend Analysis method (MDTA); (3) application of multi-scale dynamic trend analysis techniques to quantify and investigate intrinsic fluctuation trend of simulation data, heart rate variability in the aging, gender and disease and stride-to-stride variability of different age groups. According to our results, the performance of the autonomic cloud computing healthcare system is able to effectively detect intrinsic change (trend) from complex daily blood pressure and heart rate, and provide full automatic operation, accurate and objective reference indices, rather than artificial visual approach to quantitative indicators and determination. For the development of a new data analysis section, MDTA method can clearly and accurately extraction the dynamic trend variation characteristics of simulation signals on different temporal and spatial, and even have the ability to analyze the dynamic trend variation characteristics of complex heart rate variability signal on different time and space, simultaneously. The visual three-dimensional color graphs and quantitative indicators can be as evaluating aging, genders and disease of complex index of the autonomic nervous system activity. In addition, the MDTA also have applied to analyze and quantify stride-to-stride variability of different age groups, the visual three-dimensional color graphs and quantitative indicators can be as evaluating stride-to-stride variability of different age of stability index. However, the development of analysis technologies and validation of clinical not only provide new analytical technologies, applications and new insights in cardiac autonomic regulation and gait development process, even for the current wearable physiological sensing device products, cloud healthcare system and internet of things medical wisdom industry to bring new ideas, applications and requirements.