Razvoj aplikacije za določevanje srčnega utripa z digitalno kamero

magistrsko delo magistrskega študijskega programa II. stopnje Strojništvo

Žiga Pisar (Author), Primož Podržaj (Mentor)

Abstract

Človeško telo z vsakim srčnim ciklom črpa kri po celotnem telesu. Kontrakcija in relaksacija srčne mišice ustvarjata časovno spremenljivo in ponovljivo tlačno razliko znotraj kardiovaskularnega sistema. Brezkontaktna fotopletizmografija (rPPG) je novejša, cenovno ugodna in neinvazivna optična metoda zaznavanja volumetričnih krvnih sprememb po telesu. Ker človeška kri zelo dobro absorbira določen del spektra bele svetlobe, lahko s primernim senzorjem zaznamo in spremljamo majhne, očem nevidne spremembe v intenziteti absorbirane ali odbite svetlobe. V primerjavi z elektrokardiografijo in konvencionalnimi fotopletizmografi, ki zahtevajo fizičen stik, zazna rPPG z uporabo oddaljene potrošniške digitalne kamere intenziteto odbite svetlobe na koži obraza ali drugje. S procesiranjem dobljenih podatkov lahko določimo pulzni signal, posredno utrip srca, in izdelamo fiziološko sliko pacienta. Z ozirom na novi, zelo nalezljivi koronavirus, ki je povzročil globalno zdravstveno krizo, smo razvili mobilni robot za brezkontaktno določevanje utripa srca, podprtega s sistemom za oddaljeno nastavljanje ostrine. Mobilni robot smo nadgradili z namensko strojno opremo in izvedli eksperiment v treh svetlobnih pogojih: v notranjih prostorih, zunaj in pod studijsko osvetlitvijo. Meritve so pokazale, da lahko že v sobnih pogojih, s pomočjo algoritma rPPG, izmerimo frekvenco srca z napako, manjšo od 1 BPM. Aplikacija mobilnega robota z rPPG ni samo dobra alternativa konvencionalnim merilnikom pulza, temveč v časih, ko so fizični kontakti močno odsvetovani, nudi tudi možnost oddaljenega spremljanja utripa srca z nizkocenovno kamero in namenskim vgradnim sistemom.

Keywords

magistrske naloge;fotopletizmografija;mobilni robot;rPPG;Raspberry Pi;frekvenca srca;

Data

Language:	Slovenian
Year of publishing:	2021
Typology:	2.09 - Master's Thesis
Organization:	UL FS - Faculty of Mechanical Engineering
Publisher:	[Ž. Pisar]
UDC:	612.172:004.4:007.52(043.2)
COBISS:	78562051
Views:	424
Downloads:	41
Average score:	0 (0 votes)
Metadata:

Other data

Secondary language:	English
Secondary title:	Development of an application for pulse monitoring using a digital camera
Secondary abstract:	Every cardiac cycle of the human body comprises of two periods of activity: relaxation and contraction of the heart muscle. With the first, the human heart relaxes and refills with blood (diastole), followed by a robust contraction (systole) that pumps the fluid throughout the cardiovascular system. This rhythm causes a time dependent pressure pulse that drives a continuously changing volumetric flow, which can be detected due to the blood`s good light absorbance. Remote photoplethysmography (rPPG) is a novel, non-invasive and cost-effective method of detecting changes in blood volume by illuminating the dermal tissue and sensing infinitesimal intensity changes. Compared to conventional electrocardiography, rPPG does not require physical contact to determine a human`s heart rate. Using an ordinary digital camera, with a suitable rPPG algorithm, one can indirectly determine the heart rate by first measuring the pulse wave of the flowing blood. This way, an important physiology metric is obtained with information on the subjects` health. With respect to the novel and highly contagious Coronavirus that evolved into a global health crisis, we have developed and tested a telemedicine application using a tele-operated mobile robot, outfitted with a high resolution camera. An experiment was conducted in three scenarios: living room lighting, direct sunlight and in a photography studio setting. In the case of former, the average error measured was less than 1 BPM. The gathered results clearly show that a rPPG algorithm on a mobile platform works not only as a functional alternative to other conventional means of determining pulse rate, but also as a novel concept capable of decreasing the healthcare system`s workload.
Secondary keywords:	master thesis;photoplethysmography;mobile robot;rPPG;Raspberry Pi;pulse rate;
Type (COBISS):	Master's thesis/paper
Study programme:	0
Embargo end date (OpenAIRE):	1970-01-01
Thesis comment:	Univ. Ljubljana, Fak. za strojništvo
Pages:	XXIV, 65 str.
ID:	13366721