Razgradnja optičnih interferenčnih in inercijskih signalov za analizo človekovih vitalnih funkcij

Šprager, Sebastijan

| | SLO | ENG | Cookies and privacy

Bigger font | Smaller font

First page > Show document

Show document

Title:	Razgradnja optičnih interferenčnih in inercijskih signalov za analizo človekovih vitalnih funkcij
Authors:	ID Šprager, Sebastijan (Author) ID Zazula, Damjan (Mentor) More about this mentor... ID Đonlagić, Denis (Comentor)
Files:	DR_Sprager_Sebastijan_2013.pdf (9,56 MB) MD5: 1C3602B65A541BF4C9B8173295CC2B50 PID: 20.500.12556/dkum/d33cfa72-3340-4b12-8ad4-2c0b3fe78a2a
Language:	Slovenian
Work type:	Dissertation
Typology:	2.08 - Doctoral Dissertation
Organization:	FERI - Faculty of Electrical Engineering and Computer Science
Abstract:	V doktorski disertaciji raziskujemo nove pristope, ki za nemoteče spremljanje človekovih vitalnih funkcij temeljijo na razgradnji optičnih interferenčnih in inercijskih signalov. Uporabljamo optični interferometer. Gre za izredno občutljiv senzor, ki je ob posrednem ali neposrednem stiku s človeškim telesom sposoben zaznati še tako majhne perturbacije, povzročene z mehanskimi in akustičnimi vplivi človekovih vitalnih funkcij. Njihovi prispevki so superponirani v interferenčnem signalu. Uvajamo nove dekompozicijske postopke, ki iz takega sestavljenega signala izluščijo prispevke posameznih opazovanih vitalnih funkcij. Inercijske signale, ki opredeljujejo človekovo gibanje, zajemamo s pospeškometrom. Izmerjeni pospeški so primerni za analizo in identifikacijo hoje. Analizni postopek poudarja ciklostacionarne lastnosti hoje s pomočjo statistik višjih redov. V začetnem poglavju pregledamo trenutno stanje tehnike na področju nemotečega spremljanja in analize vitalnih funkcij. Razložimo fiziološke značilnosti, ki so pomembne pri spremljanju človekovih vitalnih znakov. Razpoznavamo jih z razvitimi algoritmi iz nemotečih meritev, hkrati pa jih uporabljamo za referenco. Posvečamo se predvsem srčnemu utripu, dihanju in gibanju. V posebnem poglavju podrobneje predstavimo tudi optični interferometer, ki ima kot senzor najpomembnejšo vlogo pri naših raziskavah. Pri snovanju metodološkega aparata smo morali najprej razviti različne pristope za demodulacijo interferenčnih signalov. Ti so namreč frekvenčno modulirani, zato pomeni demodulacija uvodni korak v njihovo razgradnjo. Metode za razgradnjo optičnih interferenčnih signalov temeljijo na različnih fizioloških značilnosti vitalnih funkcij, katerih energijska vsebina prevladuje v različnih frekvenčnih pasovih. Pristopi, ki smo jih razvili in raziskali, temeljijo na skupinah filtrov, časovno-frekvenčni analizi, časovno-merilni analizi, nelinearnem razširjanju in večkanalni dekompoziciji, nevronskih mrežah ter nazadnje še večmetodnem pristopu. Točnost razpoznavanja srčnih utripov dodatno izboljšamo z optimiziranim določanjem njihovih pojavljanj v času. Optimizacijo opravimo s statističnima analizama dvodimenzionalnih histogramov in največje izkustvene verjetnosti. Predstavimo tudi nov postopek za analizo hoje, ki temelji na statistikah višjih redov in je sposoben razpoznavati različne osebe in načine hoje, hkrati pa sklepati o njihovi medsebojni podobnosti. Uspešnost razvitih metod ovrednotimo z več eksperimenti. Optični interferometer uporabljamo kot posteljni in telesni senzor, meritve pa smo izvedli v nadzorovanih laboratorijskih pogojih. Sledili smo dvema protokoloma: opazovanci so mirovali ali pa so bili telesno aktivni, tako da smo dosegli spremenljiv pulz. S pospeškometri smo merili tudi parametre hoje. Potrdili smo, ali je možno samo iz pospeškov pri hoji ugotavljati identiteto opazovancev in razločevati med njihovimi različnimi načini hoje, pa tudi, kakšen vpliv imajo na hojo različne trdne podlage. Rezultati, ki smo jih dobili z ovrednotenjem eksperimentalnih podatkov, so pokazali visoko učinkovitost in točnost. Predlagani pristopi za razgradnjo optičnih interferenčnih signalov so povsem primerljivi z rezultati obstoječih metod za nemoteče spremljanje vitalnih funkcij ali pa jih celo presegajo. Podobno pokažemo tudi za postopek, ki analizira in identificira hojo.
Keywords:	biomedicinska tehnika, obdelava signalov, sestavljeni biomedicinski signali, posteljni senzor, telesni senzor, optični senzor, interferometrija, pospeškometer, človekove vitalne funkcije, razpoznavanje srčnega utripa, razpoznavanje dihanja, analiza hoje, balistokardiogram, fonokardiogram, elektrokardiogram, filtri, časovno-frekvenčna analiza, večločljivostna analiza, indeks aktivnosti, kompenzacija konvolucijskih jeder, nevronske mreže, večmetodni pristop, največja izkustvena verjetnost, statisti
Place of publishing:	Maribor
Publisher:	[S. Šprager]
Year of publishing:	2013
PID:	20.500.12556/DKUM-40108
UDC:	004.93:004.98(043.3)
COBISS.SI-ID:	266655488
NUK URN:	URN:SI:UM:DK:17QCKSAR
Publication date in DKUM:	12.04.2013
Views:	2664
Downloads:	281
Metadata:
Categories:	KTFMB - FERI
:	Copy citation

Average score:	(0 votes)
Your score:	Voting is allowed only for logged in users.
Share:

Hover the mouse pointer over a document title to show the abstract or click on the title to get all document metadata.

Secondary language

Language:	English
Title:	Decomposition of optical interference and inertial signals for analysis of human vital functions
Abstract:	In this doctoral dissertation we investigate new approaches for unobtrusive monitoring of human vital functions, based on the decomposition of optical interference and inertial signals. We use optical interferometer. It is an extremely sensitive sensor. When in direct or indirect contact with human body, sensor is capable of detecting small perturbations caused by mechanical and audible effects of human vital functions. Their contributions are superimposed in the interference signal. We introduce new decomposition approaches that are able to extract the contributions of each observed vital function from such compound signal. Inertial signals caused by human movement are acquired by accelerometers. The acquired accelerations are suitable for the gait identification and analysis. The analysis procedure emphasises cyclostationary gait features by using higher-order statistics. Introductory chapter surveys the state-of-the-art in the area of unobtrusive monitoring and analysis of human vital signs. We explain the physiological characteristics that are important when performing monitoring of human vital signs. The detection of vital signs is performed by using developed algorithms from unobtrusive measurements. To assess the algorthms’ efficiency, we observe in parallel standard referential signals. Our attention is paid to the heartbeat, respiration and movement. A special chapter introduces optical interferometers their important role in our research. While designing the methodological apparatus, we had to develop different approaches for the demodulation of the interferometric signals. These are frequency modulated, hence a demodulation step is first necessary for their decomposition. The decomposition methods for optical interferometirc signals are based on different physiological characteristics of vital functions, whose energy content dominates in different frequency bands. Approaches that have been developed and investigated in our research deploy filter banks, time-frequency analysis, multiresolution analysis, non-linear extension and multichannel decomposition, neural networks and, finally, multimethod approaches. The heartbeat detection accuracy is further improved by optimized positioning of heartbeats in time. Optimization is performed with statistical analysis by using two-dimensional histograms and maximum a-posteriori probability. We also introduce a new method for gait analysis, which is based on higher-order statistics and is capable of recognising different subjects and their walking types. Biometric and health conditions can also be inferred on their mutual similarity. The performance of developed methods is evaluated in several experiments. Optical interferometer is used as a bed sensor and body sensor. Measurements were performed in the controlled laboratory conditions. We introduced two protocols: during subjects’ rest or after their physical activity which resulted in variable heart rate. Gait parameters were also measured by using accelerometers. We investigated whether it is possible to determine the subjects’ identity and distinguish between their different walking types, as well as determined the impact of different walking surfaces. The results obtained by assessing the experimental data have shown high efficiency and accuracy. The proposed approaches for decomposition of optical interferometric signals are completely comparable to the results reproted on other known methods for unobtrusive monitoring of vital functions, or even better. Similar conclusions can be made for the proposed gait identification and analysis procedure.
Keywords:	biomedical engineering, signal processing, compound biomedical signal, bed sensor, body sensor, optical sensor, interferometry, accelerometer, human vital function, heartbeat detection, respiration detection, gait analysis, ballistocardiogram, phonocardiogram, electrocardiogram, filters, time-frequency analysis, multiresolution analysis, activity index, convolution kernel compensation, neural network, multimethod approach, maximum a-posteriori probability, high-order statistics, cumulants

Comments

Leave comment

You must log in to leave a comment.

Comments (0)

0 - 0 / 0

There are no comments!

Back