ISRN UTH-INGUTB-EX-E-2020/004-SE
Examensarbete 15 hp Juni 2020
Investigation of ECG
electrodes for burn wounds
Linus Falk
Populärvetenskaplig sammanfattning
Stora och svåra brännskador påverkar inte enbart huden utan även många system i kroppen. Brännskadevård är därför en specialistgren inom sjukvården och Akademiska sjukhuset i Uppsala var det första sjukhuset i Sverige med en avdelning specialiserad på dessa skador. De har idag tillsammans med Linköping det nationella uppdraget att ta hand om svåra brännskador.
Eftersom brännskadan påverkar många system i kroppen är det viktigt att kunna övervaka patientens parametrar under vårdtiden. Med ett EKG kan man ställa diagnos och övervaka hjärtats funktion och det används därför regelbundet inom intensivvården.
EKGt tas vanligtvis med engångselektroder som sätts på huden på standardiserade platser och tillsammans med en EKG apparat kan man registrera hjärtats elektriska aktivitet och ställa diagnoser.
På brännskadecentrum i Uppsala har man under långt tid haft återkommande problem med mycket störningar vid EKG-mätningar vilket försvårar arbetet att ställa diagnoser och övervaka patientens tillstånd. Målet med detta arbete var att granska EKG-kurvor från brännskadecentrum med uppenbara störningar och undersöka brännskadans effekt på elektroderna för att kunna rekommendera en typ av elektroder för denna typ av patienter.
Arbetet utfördes med en litteraturöversikt över EKG-instrumentering, vanliga
störningar, elektroder och hur brännskadan kunde efterliknas i ett standardiserat test av elektrodernas elektriska egenskaper. Tester gjordes sedan på elektroderna med och utan denna efterliknelse för att se hur mycket elektroderna påverkades och om det var tillräckligt för att åstadkomma de problem som avdelningen ofta har.
Resultatet blev en rekommendation av elektroder av våt gel typ eftersom de uppvisade bäst elektriska egenskaper för att minska förvrängning av EKG signalen och att risken för störningar på grund av obalans i impedans mellan elektroder på brännskada och hel hud är lägre med den typen av elektroder.
4
Acknowledgments
Many people have been kind to help and guide me through this work and with the fear of leaving someone out by mistake, I want to give a special thanks to the involved institutions:
Burn Center – Uppsala University Hospital
The section for medical technique – Uppsala University Hospital Microwaves in Medical Engineering Group – Uppsala University
Signal and system – Uppsala University
5
Table of contents
Nomenclature _______________________________________________________________ 7
1 Introduction _______________________________________________________________ 8 1.1 Background _________________________________________________________ 8 1.2 Purpose ____________________________________________________________ 9 1.3 Delimitations _________________________________________________________ 9 1.4 Method _____________________________________________________________ 9 1.5 Objective ___________________________________________________________ 9
2 Theory ___________________________________________________________________ 10 2.1 Skin and burn wounds ________________________________________________ 10 2.1.1 Normal skin __________________________________________________ 10 2.1.2 Burns and severity classification __________________________________ 10 2.1.3 Care and treatment ____________________________________________ 11 2.2 Heart and ECG measurements _________________________________________ 11 2.3 ECG electrodes _____________________________________________________ 15 2.3.1 ECG electrode introduction ______________________________________ 15 2.3.2 Equivalent circuit of an ECG electrode _____________________________ 18 2.3.3 Equivalent circuit of an ECG electrode placed on the skin ______________ 21 2.3.4 ECG electrode quality control according to ANSI/AAMI ________________ 22 2.4 ECG measuring technique and common artefacts __________________________ 22
3 Method __________________________________________________________________ 25 3.1 Devices and material _________________________________________________ 25 3.1.1 Data acquisition (DAQ) device ___________________________________ 25 3.1.2 Graphical programming language: ________________________________ 25 3.1.3 Electrodes ___________________________________________________ 25 3.1.4 Ringer’s acetate ______________________________________________ 26 3.2 Investigation of artefacts in ECG signals __________________________________ 26 3.3 Burn wound simulation ________________________________________________ 26 3.4 Electrode selections and measurements __________________________________ 27 3.4.1 DC offset ____________________________________________________ 27 3.4.2 10Hz AC impedance ___________________________________________ 28 3.4.3 Adhesiveness ________________________________________________ 29 4 Result and discussion _____________________________________________________ 30 4.1 Artefacts in ECG signals and their counter ________________________________ 30
6
4.2 Measurements ______________________________________________________ 30 4.3 Adhesive __________________________________________________________ 33 6 Conclusion and further work ________________________________________________ 35 References ________________________________________________________________ 36
7
Nomenclature
AAMI Association for the Advancement of Medical Instrumentation ANSI American National Standards Institute
aVF Augmented vector foot aVL Augmented vector left aVR Augmented vector right
CMRR Common Mode Rejection Rate ECG Electrocardiography
Interstitium Space between the cells
QRS Graphical deflections in ECG waveform SSCD Skin surface conductance density
8
1 Introduction
In this chapter, the project works background, purpose, delimitations and method will be explained.
1.1 Background
Advanced treatment of burn victims has been carried out at Uppsala University Hospital for over 60 years and together with Linköping university Hospital they got the total responsibility for treatment of severe burns in Sweden. Since May 2011 there is a modern centre: Burn Center, for burn treatments on Uppsala University Hospital.
Large and severe burn injuries cause the human body to lose large amount of liquid through the damaged areas. This liquid leakage combined with the added liquid from
“fluid treatment” makes this type of patient difficult to use conventional “stick on” ECG electrodes on. The ECG electrodes are placed on the body in standard places and together with an ECG device the electrical activity in the heart can be picked up. In Fig.
1.1 a part of a 12-lead ECG is shown taken on one of the patients in the Burn Center.
The ECG waveforms from different leads show different amount of interference that are frequently encountered on ECG waveforms at from this category of patients, because burn wounded skin affect the performance of ECG electrodes.
Fig. 1.1 ECG waveform from burn patient
9
1.2 Purpose
The purpose of this project work is to investigate ECG-electrodes to determine which type is most suitable for extracting the ECG signal from burn wounded skin.
1.3 Delimitations
The project work is limited to investigate the properties of ECG electrodes used for measuring ECG signals from full thickness burn wounded skin. Electrodes that are not suitable for easy sterilization or not of disposable type will be excluded from the work due to hygienic reasons. No new methods to replace the conventional placement of the electrodes will be developed or investigated in this project.
1.4 Method
The project work will begin with a literature review of the basics of an ECG and theoretical and empirical methods of recreating the surface and electrical properties of burn wounded skin in an artificial way. To test the electrodes’ electrical quality an industrial standards test will be done on the selected electrodes, one reference test without burn wound replication and one with. A test of the adhesive of the electrodes will also be performed. Selecting suitable electrodes for testing will be done by consulting personnel of the Uppsala University Hospital Burn Center.
1.5 Objective
The objective of this project work is to deliver a recommendation of what sort of electrode is suitable to use on burn wounds. Following secondary objectives shall be fulfilled during the work:
• Collect and validate ECG waveforms at the Burn Center of Uppsala University Hospital
• Describe the electrical properties of burn wounded skin and how it could be replicated
• Construct a test for ECG electrodes and test them
• Validate the result.
10
2 Theory
In this chapter the relevant theory for the project work will be presented.
2.1 Skin and burn wounds
The skin is the largest organ of the human body and serves several important functions for the human body.
2.1.1 Normal skin
The skin act as a barrier against the surrounding environment and some of its main tasks are:
- protect against mechanical and chemical impact - protect against dehydration
- stop microorganism entering the body
Skin consists of two layers, the Epidermis and Dermis. Epidermis is outer layer and is around 0.1-1mm thick depending on how much wear the skin is subjected to. There are no blood vessels in the epidermis, all nutrients are transported by diffusion. The outermost layer of the Epidermis is called the Corneum stratum and consists of cornified dead cells. Dermis is the layer underneath the Epidermis and is 0.3-3mm thick. It consists of fibrous connective tissue, sweat glands, hair follicle, sebaceous glands, and a lot of blood vessels. Dermis merges gradually to the subcutaneous layer of loose fibrous connective tissue and varying amount of fat. The blood vessels in the epidermis is supplied with blood from larger vessels in the subcutaneous layer [1].
2.1.2 Burns and severity classification
Burn wounds separate themselves from many other wounds. Even though it is only one organ involved in the damage it affects almost all systems in the whole body. Burn wound treatment and care is there for a medical speciality.
Burn wounds can be caused by many things but the categorisation of the burn wounded skin is the same for all of them [2]. The skin is divided into 3 parts: the utmost part is called the epidermis followed by the dermis and the subcutaneous layers. Burn wounds are divided into categories depending on the depth of the wound.
11
Superficial, epidermal or first-degree burn has only damaged the outer layer of the skin, epidermis. Common damages of superficial burns are a typical sunburn with redness and mild swelling [3].
Superficial dermal or second-degree (2a) burn is damage to the epidermis and the superficial part of the dermis. Blisters, red and moist wound surface.
Deep dermal or second-degree (2b) burn has damaged deeper into the dermis with grey white or red often dry wound surface. Often is surgery needed to remove the damaged epidermis.
Full thickness burn has damaged all of the dermis down to the subcutaneous layers of fat and possibly also deeper tissue layers. Surgery is mandatory for healing unless the wound is very small, approximately 1cm2 [3].
In response to a burn wound an inflammatory response is triggered and blood components is leaked out from the intravascular space into the interstitium causing oedema in the damaged areas and often also in the whole body [4]. The blood components consist of blood cells, proteins and plasma. In the blood plasma ions is the main component of the solved substances with natrium being the one of highest concentrations [5].
2.1.3 Care and treatment
To replace the intravascular fluid that is leaked out from the damaged areas, fluid treatment is started with Ringer’s acetate [4]. Ringer’s acetate is an isotonic infusion liquid that does not change the volumes of the cells when injected. It contains all ions normally found in the extracellular liquid in similar concentrations [6].
2.2 Heart and ECG measurements
The heart is a muscle that consists of four chambers, left and right ventricles and atriums. The purpose of the heart is to circulate the blood in the body to deliver nutrition, oxygen and remove waste products from cells. By contracting in a specific order, shown in Fig. 2.1, it pumps the blood through the heart and out into the circulatory systems [7].
12
Fig. 2.1 Contraction of the atrium and chambers [8]
The contraction of a single muscle fibre relates to the change in potential on the surface of the muscle cells. In rest the muscle cell is polarized so that the inside of the cells has a negative charge compared to the surrounding membrane. This potential of around - 90mV difference is maintained by active transportation of Na+ through the cell membrane out of the cell. When a contraction of a muscle fibre occurs an action potential is triggered which is caused by a sudden change of the cell membranes permeability of Na+ ions flipping the potential of the membrane. A dipole field contraction wave is created in the moment when depolarization occurs traveling along the muscle fibre. It is later followed by a repolarization wave in the opposite direction [9].
These potentials caused by individual muscle fibres contractions can be picked up by electrodes connected to an ECG device and are added up to a waveform, the ECG.
These potentials are in the magnitude of 0,5 to 4 mV and in the frequency range: 0.01 to 250 Hz [10]. The waveform is shown in Fig. 2.2 where the P wave is representing the depolarization of the atria and QRS interval or QRS complex represent the depolarization of the ventricles and the T wave representing the repolarization of the ventricles [8].
13
Fig. 2.2 Intervals of the ECG [8]
For diagnosing heart problems, a 12-lead ECG is often used. The 12-leads imply that there are 12 different waveforms formed by electrodes placed in a standardized placement on the body [11]. The leads are of three types:
Bipolar extremity leads – standard leads I, II and III: In the standard leads the potential between the points shown in Fig. 2.3 is measured.
Unipolar extremity leads – aVR, aVL, and aVF: Analysis of the ECG signal from the extremities can be made easier by registering them in relationship to a point whose potential does not change during the heart cycle. This is achieved by connecting the two other extremity leads with two equally large resistors.
Fig. 2.3 The bipolar and unipolar extremity leads [12]
14
Unipolar chest leads – V1,V2, … Vn; To get more detailed information of the changes in potential in the heart, electrodes are placed around the chest in anatomicallyspecified places as seen in Fig. 2.4. The potentials in these sites are measured in relationship with a constructed point that connects the left arm, right arm, and left leg with three equal resistors.
Fig. 2.4 The unipolar chest lead placements [13]
These placements of the electrodes make it possible to observe the electrical activity in the heart from different angles [11], resulting in different shapes of the ECG waveform seen in Fig. 2.5.
15
Fig. 2.5 The 12-lead ECG waveforms [14]
2.3 ECG electrodes
In this chapter the ECG electrode is introduced, and its electrical properties explained.
2.3.1 ECG electrode introduction
ECG electrodes exist in three main categories:
• Surface electrodes
• Monopolar electrodes
• Concentric electrodes
The most commonly used ECG electrode is the disposable surface electrodes which are often covered with a thin layer of electrode-paste containing electrolytes to improve conductivity. These electrodes are kept in place by either suction, tape, glue or a strap.
It’s the potential difference between two electrodes or between one electrode and a constructed point with a ground reference electrode that are picked up and conducted to the ECG device that filters and amplifies the signal and displays it for interpretation.
The sum of the electrical activity in the tissue under the electrode is picked up by
16
surface electrodes and they are therefore not suitable when locating the exact position of the source of a signal.
ECG surface electrodes job is to convert the current in the body that consists of ion transport to electrons through the surface of the electrode, often combined with an electrolyte [15]. Surface ECG electrodes can be further categorised as either polarizable or non-nonpolarizable [16]. The most widely used non-polarizable disposable electrode is the silver chloride Ag-AgCl electrode because of its property to easily exchange ions with its surroundings. The easy exchange of ions makes it not as polarized as other conductive materials like stainless steel or platinum. Polarization of the electrode resembles the surfaces of a capacitor with charges distributed on two sides, the electrode surface as one side and the electrolyte the other. This double layer is called the Helmholtz double layer and its simplest form is shown in Fig. 2.6. Polarization can in some cases lead to difficulties to register an ECG signal in the low frequencies or even build up a voltage so high that it blocks the input of the amplifier in the ECG device [15].
Fig. 2.6 The Helmholtz layer [17]
The Ag-AgCl electrodes are often made with either a solid or wet gel for the electrolyte.
The solid gel or hydrogel is held together by crosslinked polymers that can by absorption contain more than 99% water. This solid gel can either just contain the electrolyte or the adhesive for the electrode also. The wet gel is in a liquid state that lowers the resistance in combination with the skin by penetrating the outer layer of it [18]. One of the wet gel electrodes design features is that it reduces the risk of artefacts in the ECG curve by having a buffer layer of isotonic electrolyte between the surface of
17
the patient and the electrode, shown in Fig. 2.7. This layer absorbs the movement of the electrode in relationship to the patient. This is to maintain a constant polarization or half-cell potential [17].
Fig. 2.7 Buffer layer in wet gel ECG electrode [17]
Comparing the low frequency conductivity between solid and wet gel electrodes on skin, shown in Fig. 2.8 the wet electrode (b) performed up to 8,5 times better when the patient is resting/passive. When the patient was exercising the conductivity improves for the solid gel (a) in contrast to the wet gel electrode that performed worse. The reason for this is probably that the sweat improves the conductivity in the skin solid gel interface but decreased it in the wet gel case because the wet gel was replaced by sweat that is less conductive then the gel. Solid gel as contact medium to skin shows in general more capacitive coupling and performs there for worse in low frequency applications. When choosing wet or solid gel electrode these characteristics should be considered [18].
18
Fig 2.8 Conductivity with skin of a solid gel (a) and wet gel (b) electrode [18].
2.3.2 Equivalent circuit of an ECG electrode
The ECG signal is an AC signal and the electrical property of the electrode is therefore described as an impedance. Impedance Z is complexed valued, consisting of resistance R as its real part and reactance X as its imaginary part. Mathematically, the impedance can be expressed as:
𝑍 = 𝑅 + 𝑗𝑋 (2.1)
Where the reactance can be either capacitive or inductive. The capacitive reactance is described as negative in the imaginary plane and inductance positive. The impedance can also be described in exponential form where the magnitude is the hypotenuse of the resistance and reactance together with the angle θ between them. This is denoted in the form:
𝑍 = |𝑍|𝑒𝑗𝜃 = √𝑅2+ 𝑋2𝑒𝑗𝜃 (2.2)
𝜃 = tan−1𝑋
𝑅 (2.3)
19
If the impedance is purely capacitive its angle is -90 degrees or if purely inductive + 90 degrees. With Ohm’s law the impedance can be calculated from the complexed valued voltage V and current I, shown in Eq. (2.4) [19].
𝑍 =𝑉
𝐼 (2.4)
The impedance of a surface electrode typically has resistance of <10kOhm and capacitance <0.1µF [15]. The current through a capacitor is described by Eq. (2.5). That means that the voltage can’t change instantaneously over a capacitor because that would need a current that is infinitely large. The voltage vc over capacitor in a RC circuit will therefore lag after the voltage over the resistor vr as shown in Fig. 2.9. This is called the phase shift and is measured in degrees or radians [19].
𝐼(𝑡) = 𝐶𝑑𝑉(𝑡)
𝑑𝑡 (2.5)
Fig. 2.9 Phase shift between VR and VC [19]
The value of an unknown impedance Z often needs to be determined. This can be done by connecting an AC source and a known resistance Rref with it in series as shown in fig. 2.10. The value of Z can be calculated in the following manner: [20].
𝑍 = |𝑍|𝑒𝑗𝛼 = 𝑉2𝑅𝑟𝑒𝑓
√𝑉12− 2𝑉1𝑉2cos 𝜃 + 𝑉22 𝑒𝑗𝛼 (2.6)
20
𝛼 = 𝜃 − tan−1 −𝑉2sin 𝜃
𝑉1− 𝑉2cos 𝜃 (2.7)
Where V1 and V2 is the amplitude over the known resistance and the unknown impedance, respectively, and θ is the phase shift.
Fig. 2.10 Circuit for determining value of an unknown impedance Z
The disposable ECG electrode works by the principle of half-cell voltage. This voltage is created whenever an ionic solution comes in contact with an electrode metal and the electrode metal tries to exchange ions with the electrolyte [15]. In the case of non- polarizable electrode, the current between the metal and the electrolyte passes through the electrolyte – electrode interface. This is possible by the oxidation of the electrode that forms cations and electrons, the cations travel out in the electrolyte while the electron is carried through the lead wire. In the electrolyte the anions are traveling towards the electrode to deliver electrons to the electrode. The uneven distribution of these cations and anions form the half-cell voltage and act also as a polarized interface.
In contrast the polarized electrode does not allow a current to pass freely through the electrode – electrolyte interface, instead the interface acts like a capacitor where the currents are displaced.
21
The equivalent circuit for the non-polarizable electrode can therefore be described in Fig. 2.11. The resistor Rd in parallel represents the electrical resistance of the current that passes through the electrolyte with the help of oxidation. Because the non- polarizable electrode is not perfect, a polarization occurs that is represented by the capacitance Cd. The resistance Rs represents the resistance in the electrolyte. The half- cell voltage is represented as a battery, Ehc [21].
Fig. 2.11 Equivalent circuit model of ECG electrode [22]
2.3.3 Equivalent circuit of an ECG electrode placed on the skin
Epidermis is the part of the skin that stands for most of the impedance and it vary a lot between patients, a Re of around 300kOhm for 10Hz signal is not unusual, shown in the simplified electrical model in Fig. 2.12 [17]. The capacitive properties are added because the epidermis/corneum stratum forms a semipermeable layer for ions, so there can be a difference in concentrations of ions. This difference forms a small potential difference Ese and acts as a capacitive surface Ce. The electrical properties change in the case of perspiration, and a parallel RC (Rp and Cp) circuit is added to describe the wall of the glands and the conducting sweat [22]. The Epidermis impedance can in normal cases be lowered by carefully abrading the skin to remove the utmost layer, the stratum corneum. The capacitive The impedance of dermis and subcutaneous layers are often described with only a resistance Ru of around 100Ohm [18].
22
Fig. 1.12 The electrical properties of skin [22]
2.3.4 ECG electrode quality control according to ANSI/AAMI
Disposable ECG electrodes can be characterized by using a series of tests developed by the Association for the Advancement of Medical Instrumentation (AAMI) which is an accredited standards development organization by the American National Standards Institute (ANSI). This test is to ensure safety and efficiency in the use of the electrodes in clinical use. A selection of the standard is presented in the list below [23].
1. DC offset voltage: A pair of electrodes that are connected gel to gel after 1 minute of stabilization time shall not exceed 100mV offset voltage.
2. Average value of 10 Hz impedance: The average value of the impedance of 12 pairs of electrodes connected gel to gel, max 2kOhm, individual pair max 3kOhm. The current when testing should not exceed 100µA.
2.4 ECG measuring technique and common artefacts
One of the difficulties in measuring the ECG signal is to limit the interference caused by the capacitive connection between the patient and the main 230V 50Hz network. This is often solved by differential amplifiers that have a high Common Mode Rejection Rate (CMRR) [24]. With high CMRR the interference that is present at both inputs of the amplifiers is cancelled out. This is only possible if the common mode signal (the
23
interference) stays at the same amplitude and phase at both inputs. Imbalance in impedance between the electrodes (Zex in Fig. 2.13) can make the common mode signal become differential by voltage dividing over the electrode impedance, the amplifier can in these cases not cancel out all the interference [25]. Common mode to differential mode conversion can also occur because of voltage drop in the body (Ztx in Fig 2.13) between the measuring electrodes from the coupled 50Hz current [26].
Fig. 2.13 Capacitive coupling to main 230V, 50 Hz network [27]
Artefacts in an ECG makes it difficult to interpret and to set a correct diagnosis. Some common artefacts in ECGs are described and explained bellow:
Wandering baseline in Fig. 2.14, is a low frequency artefact well below 1Hz. This wandering baseline makes it very difficult to interpret the S-T interval of the ECG.
Wandering baseline is often caused by perspiration, respiration, patient movement and poor electrode contact [28].
24
Fig. 2.14 Wandering baseline [28]
Muscle tremor artefact in Fig. 2.15 is the result of muscle activity from other muscles than the heart. These are best avoided by trying to keep the patient warm to avoid shivering and tell them to be still and relaxed [29].
Fig. 2.15 Muscle tremor artefact [28]
AC 50-60Hz interference in Fig. 2.16 is an artefact that makes the baseline thick and fuzzy. It is often related to poor electrode contact, dried electrode gel from incorrect storage and defective cables [26].
Fig. 2.16 The 50-60Hz interference [28]
25
3 Method
In this chapter the method to achieve the objectives in this work is described. To find what type of disposable electrode is best suitable for burn wounds the following issues are investigated: the ECG-waveforms from Burn Center and the effect of the electrical properties and adhesive of the electrode by the burn wound.
3.1 Devices and material
In this section the devices and material used in this project are introduced.
3.1.1 Data acquisition (DAQ) device
A National Instruments USB-6210 data acquisition device was used for the measurements. USB-6210 is a multifunctional 16-Bit, 250kS/s DAQ that offers 16 analog inputs and 4 digital inputs [30].
3.1.2 Graphical programming language:
Laboratory Virtual Instrument Engineering Workbench (LabVIEW) is a graphical programming language that was used to collect and calculate the measurements from the DAQ. A LabVIEW program is called a virtual instruments (VI) and consists of a block diagram, front panel and a connector pane [31]. Collection and calculations of data from the DAQ was programmed in the block diagram and the results was displayed in the front panel.
3.1.3 Electrodes
In table 3.1 the selected electrodes for the test are presented. The notes describe specification of use for the electrode.
26
Table 3.1 Selected electrodes
Manufacturer Model Gel Notes
Ambu Bluesensor L-00-S/25 Wet
For long time ECG
3M 2670-5 Solid Repositionable
Ambu Bluesensor R-00-S/25 Wet Exercise test
Milmedtek T-VO01 Wet Dry skin
Medtronic Arbo Solid
For X-ray and MRI
Ambu Whitesensor WSP30-00-S/50 Solid Exercise test 3.1.4 Ringer’s acetate
Ringer’s acetate is an isotonic infusion liquid. It contains all ions normally found in the extracellular liquid in similar concentrations [6].
3.2 Investigation of artefacts in ECG signals
By comparing the artefacts found in the two ECG waveforms received from Burn Center with common artefacts, possible sources could be identified.
3.3 Burn wound simulation
The main difference of the measurement of an ECG on a burn wounded patient is the often lack of the outer part of the skin, the epidermis, and the presence of extracellular liquid leaking from the areas where the electrode is supposed to be attached. To investigate the effect this wound has on the electrode’s electrical properties 0.5ml Ringer’s acetate, which is a liquid very similar to extra cellular liquid was placed between two electrodes gel to gel, shown in Fig. 3.1. The electrodes were then tested using the two test methods (AAMI) described in theory chapter. This test was repeated without Ringer’s acetate for comparison. The purpose of these tests was to see if and how the electrodes electrical properties changed and if there were any differences in the results between the different types of gel used on the electrodes.
27
Fig. 3.1 Applying Ringer’s acetate to electrode
3.4 Electrode selections and measurements
Selection of ECG electrodes was done consulting an intensive care nurse at Burn Center. The goal was to test a variety of disposable electrodes with different kinds of gels that were easily available from their suppliers. One electrode for dry skin was selected for comparison in the effect on the adhesive.
The two methods to test the electrodes are described here. The results from each test were collected in an excel form and saved. The results were then collected into graphs and displayed with the mean value of 12 electrode pair tests and with the standard deviation.
3.4.1 DC offset
The electrodes were connected to a 3.5mm audio cable with snap-on connectors to one of the DAQ’s differential ports by a 3.5mm port. The input configuration for the DAQ was set to Differential mode in LabVIEW and 10kOhm resistors for bias currents were placed from the differential ports to the AIGND port.
28
Fig. 3.2 DC offset test circuit 3.4.2 10Hz AC impedance
The 10Hz AC impedance was measured using a Velleman PCSU200 oscilloscope with signal generator. The signal generator was set to sine wave, 10Hz and 8V peak to peak.
The electrodes were connected in series with a 1MOhm resistor to ensure that the current was lower than the maximum value according to AAMI standard. The voltage and phase shift over the resistor (VR) and electrodes (VE) were measured in differential mode with the DAQ. The impedance was calculated in LabVIEW using Eqs. (2.6) and (2.7). The measurement was done with a sample rate of 1kHz for 1 second and then repeated 30 times, a mean value was then calculated for each pair of electrodes. This is to ensure that a single measurement error would not have a big impact on the result.
R_output shown in fig. 3.3 is the output resistance of the signal generator and does not affect the result in this test.
29
Fig. 3.3 Impedance test circuit 3.4.3 Adhesiveness
During the measurements of the electrode, notes were taken on the effect on the adhesive of the electrodes. Noted was how the liquid spread over the electrode and if the adhesive stopped working.
30
4 Result and discussion
The results from the measurements and observations are presented and discussed in this chapter.
4.1 Artefacts in ECG signals and their counter
The unidentified ECG waveforms found in the appendix and a part of it in Fig 3.3 shows a lot of 50Hz interference. Electrode related causes is dried out gel, poor contact and common mode to differential mode conversion by impedance imbalance between the electrodes.
Fig. 3.3 Part of unidentified ECG showing (a) 50Hz interference and (b) ventricle contractions
The most common solutions for this kind of problems are:
Dried out gel is a problem caused by storing the electrode wrong and is easily corrected by following the manufactures guidelines.
Poor contact is caused by badly performing adhesive and is best solved by choosing an electrode that is developed for the application in question, for an example sweaty skin or exercise ECG.
Impedance imbalance is controlled by reducing the differences in skin-electrode impedances between electrodes.
4.2 Measurements
The results of the measurements are presented here as graphs with the blue bars (test 1) representing the test without Ringer’s acetate and the red bars (test 2) with Ringer’s acetate.
31
The measurements of the electrodes showed that the impedance of the electrodes changed when Ringer’s acetate was introduced between the electrode pairs. This is expected when introducing a conductive liquid between two electrodes and the result was in almost all cases a lower impedance, shown in Fig. 4.1. A two-sided T-test with significance level of 5% showed a convincing difference between the mean value in test 1 and 2 in impedance and phase. Lower impedance can cause more imbalance in impedance between electrodes placed on burn wound and skin, which increases the risk of 50Hz interference. The change in impedance because of Ringer’s acetate was often less than 100Ohm and is very small in comparison with the impedance of dry skin of around 300kOhm if the electrodes are placed on both skin and burn wound.
Recommending an electrode that showed least change is there for not necessary. Even though the solid gel electrodes showed lower impedance should wet gel electrodes still be considered. This is because of the much higher conductivity in combination with non-sweaty skin to reduce impedance imbalance when electrodes are placed on both skin and burn wound.
Fig. 4.1 impedance of ECG electrode pairs
Increased phase shift was observed in almost all cases in test 2, shown in Fig. 4.2. This could be caused by either lower resistance between the electrodes or increased polarization because of changed concentration of ions in the electrode-gel interface. An increase in DC-offset (shown in Fig. 4.3) would have also been observed if increase in
0 100 200 300 400 500 600 700 800 900 1000
3M 2670 Medtronic arbo AMBU BLUE L- 00-S25
AMBU BLUE R- 00-S25
Milmedtek T- VO01
AMBU White WPS30-00-S/50
Impedance |Z| (Ohm)
Test 1 Test 2
32
polarization occurred, but no statistical difference (Two-sided T-test with 5%
significance level) between test 1 and 2 was found. Capacitive coupling impacts the ECG by increasing high pass filtering of the signal and is not wanted. The wet gel electrodes with general low capacitive coupling is therefore a better choice.
Fig. 4.2 Phase shift between voltage over resistor and electrode pair
Fig. 4.3 DC offset in electrode pair after 1 min stabilization time
-45 -40 -35 -30 -25 -20 -15 -10 -5 0
3M 2670 Medtronic arbo AMBU BLUE L- 00-S25
AMBU BLUE R- 00-S25
Milmedtek T- VO01
AMBU White WPS30-00-S/50
Phase (º degrees)
Test 1 Test 2
-0,5 0 0,5 1 1,5 2 2,5 3 3,5 4 4,5
3M 2670 Medtronic arbo AMBU BLUE L- 00-S25
AMBU BLUE R- 00-S25
Milmedtek T- VO01
AMBU White WPS30-00-S/50
DC Offset (mV)
Test 1 Test 2
33
The resistors for bias currents were not tested and could therefore have been better selected. The manufacturer of the DAQ, National instruments recommendation is to have resistors in the interval 10kOhm up to 100kOhm and is suitable for the environment the measurements are done [32]. In this case the resistors were 10kOhm and could have had the effect of lowering the input impedance making the measurements incorrect. This measurement problem could explain why no statistical difference was found between the tests. The result from the DC offset should therefore be interpreted carefully. Increased phase shift in test 2 that could be caused by increased polarization of the electrode needs better measurements and other test methods to be confirmed.
4.3 Adhesive
The notes of how the adhesive reacted in test 2 with Ringer acetate are presented in table 4.1 and discussed below.
Table 4.1 Notes on how the adhesive reacted to Ringer acetate
Electrode: Notes
3M 2670 Liquid not absorbed, some pushed out from the electrode surface.
Medtronic arbo Liquid not absorbed, some pushed out from the electrode surface.
AMBU BLUE L-00-S25 All liquid stayed inside electrode, absorbed by the sponge with gel, no effect on adhesive AMBU BLUE R-00-S25 All liquid stayed inside electrode, absorbed by the sponge with gel, no effect on adhesive Milmedtek T-VO01 Adhesive in the center of electrode stopped working after a couple of minutes
AMBU White WPS30-00-S/50 Liquid not absorbed, some pushed out from the electrode surface.
The volume of Ringer’s Acetate applied to the electrodes was chosen so that the smallest electrode would not flood over before the electrodes were put together, this meant that for some of the electrodes it only covered the central part of the electrode and didn’t affect the adhesive as much as the smaller electrodes. The control electrode for dry skin in the adhesive test was the only electrode that showed worse performance in test 2, with increased impedance. It was also the only electrodes that started to loosen from each other during the same test. Any conclusion except following the
34
manufacturers recommendation for use regarding the adhesive could therefore not be made in this test.
35
6 Conclusion and further work
To conclude from the results of the measurements and known differences between different types of electrodes presented in this work it would be recommended to use electrodes of wet type because of the general lower high pass filtering and its negative effect on the ECG signal. The increase in phase shift with Ringer’s Acetate with possibly higher polarization support this choice also. The lower impedance on regular skin from the wet gel decrease the risk of interference because of impedance imbalance when electrodes are placed on both skin and burn wound. Of the tested electrodes should the Ambu Bluesensor R-00-S/25 be recommended because of its wet gel and adhesive that is developed for sweaty/wet skin.
Suggestions for further investigation would be to see if the interference could be solved by impedance balancing between electrodes on dry skin and burn wound with an external impedance. Another suggestion would be to investigate if there is a greater coupling between the wet burn wounds and the main 230V 50Hz network causing higher currents and voltage drops in the body increasing the risk of common mode to differential mode conversion.
36
References
[1] O. Sand, E. Haug, J. G. Bjålie and Ø. V. Sjaastad, “Från cell till kropp,” in Människokroppen: Fysiologi Och Anatomi, Stockholm, Liber, 2007, pp. 95-100.
[2] V. Tiwari, "Burn wound: How it differs from other wounds?", Indian J Plast Surg. vol. 45no. 2, pp. 364–373, May-Aug., 2012. [Online]. Available:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3495387/. [Accessed: May 2020].
[3] F. Fridén, F. Huss och M. Furebring, "Sårbehandling," in
Brännskadekompendium, Uppsala university hospital, Uppsala, pp. 31-32, 2019.
[4] F. Fridén, F. Huss och M. Furebring, "Vätskebehandling första dygnet," in Brännskadekompendium, Uppsala university hospital, Uppsala, pp. 31-32, 2019.
[5] ”Blod” in Nationalencyklopedin [Online document], Available: NE.se, http://www.ne.se. [Accessed: May 2020]
[6] “Ringer-acetat Braun” in FASS [Online document], Available: FASS.se, www.fass.se. [Accessed: May 2020]
[7] O. Sand, E. Haug, J. G. Bjålie and Ø. V. Sjaastad, “Cirkulationssystemet,” in Människokroppen: Fysiologi Och Anatomi, Stockholm, Liber, 2007, pp. 270- 275.
[8] J. D. Enderle, S. M. Blanchard och J. D. Bronzino, ”Anatomy and Physiologi,”
in Introduction to biomedical engineering, (2nd ed), Elsivier/Academic press , p.
98.
[9] M. Lindén and P. Öberg, “Elektrokardiografi-EKG,” in Jacobsons medicin och teknik, 5th ed., Stockholm, Studentlitteratur, 2006, p. 143.
[10] M. Lindén and P. Öberg, “Elektrofysiologisk mätteknik,” in Jacobsons medicin och teknik, 5th ed., Stockholm, Studentlitteratur, 2006, p. 152.
[11] M. Lindén and P. Öberg, “Elektrokardiografi-EKG,” in Jacobsons medicin och teknik, 5th ed., Stockholm, Studentlitteratur, 2006, pp. 144-146.
[12] M. Gertsch, in The ECG manual, London, Springer, 2009, p. 6.
[13] M. Ramonó, in Text Atlas of practical electrocardiography, London, Springer, 2015, p. 10.
37
[14] M. Gertsch, “Practical Approach” in The ECG manual, London, Springer, 2009, p. 19.
[15] M. Lindén and P. Öberg, “Elektrofysiologisk mätteknik,” in Jacobsons medicin och teknik, 5th ed., Stockholm, Studentlitteratur, 2006, pp. 153-156.
[16] M. James och Z. George, ”Surface Electrodes,” i Biomedical technology and devices handbook, New york, CRC press, 2004, p. 82.
[17] R. Khandpur, ”Electrodes for ECG,” i Handbook of Biomedical instrumentation, 2nd red., New Dehli, Tata McGraw-Hill Publishing Company Limited, p. 52.
[18] S. Grimnes and Ø. G. Martinsen, “electrodes” in Bioimpedance and
Bioelectricity Basics. (2nd ed.) 2008. [Ebook] Available: www-sciencedirect- com.
[19] R. L. Boylestad, “The basic elements and phasors,” in Introductory circuit Analasys, 11th ed., Upper Saddle, New Jersey, Pearson Education Ltd, 2007, pp.
587-637.
[20] Tektronix, “Oscilloscope Measurement Lab,” 2013. [Online]. Available:
https://www.tek.com/document/how-guide/oscilloscope-measurement-lab- measuring-impedance-and-capacitance. [Accessed May 2020].
[21] S. Lee and J. Kruse, Biopotential Electrode Sensors in ECG/EEG/EMG Systems [Online]. Available: https://www.analog.com/media/en/technical-
documentation/white-papers/ECG-EEG-EMG_FINAL.pdf.
[22] J. G. Webster, “Biopotential electrodes,” in Medical instrumentation:
application and design, 4 ed., Hoboken, New Jersey, John Wiley & Sons, Inc, 2010, pp. 202-208
[23] R. Khandpur, ”Electrodes for ECG,” in Handbook of Biomedical
instrumentation, 2nd ed., New Dehli, Tata McGraw-Hill Publishing Company Limited, p. 55, 2003.
[24] M. Lindén and P. Öberg, “Elektrofysiologisk mätteknik,” in Jacobsons medicin och teknik, 5th ed., Stockholm, Studentlitteratur, 2006, pp. 156-159.
[25] B. Crone, "Common-Mode Rejection: How It Relates to ECG Subsystems and the Techniques Used to Provide Superior Performance." Analog Devices.
[Online]. Available: https://www.analog.com/en/technical-articles/common- mode-rejection.html. [Accessed: May 2020].
38
[26] J. Huhta and J. G. Webster, “60-Hz Interference in Electrocardiography,” IEEE Transactions on biomedical engineering, vol. BME-20, iss. 2, pp. 91-101, March 1973.
[27] ”Interference redcution in ECG recordings by using dual ground electrode,” in XIX IMEKO World Congress, Lisabon, Portugal, 2009.
[28] R. X. Stroobandt, S. S. Barold och A. F. Sinnaeve, ”Errors in
electrocardiography,” in ECG from Basics to Essentials, West Sussex , John Wiley & Sons, Ltd, 2016, pp. 394-395.
[29] A. Davies och A. Scott, ”Quality Issues Pertaining to ECG Recording,” in Starting to read ECGs, London, Springer, 2014, p. 42.
[30] National Instruments, "ni.com" [Online]. Available: https://www.ni.com/sv- se/support/model.usb-6210.html.
[31] National Instruments, "Getting Started with LabVIEW," [Online]. Available:
https://zone.ni.com/reference/en-XX/help/371361R- 01/lvhowto/lv_getting_started/.
[32] National Instruments, “How to select Bias resistors,” [Online]. Available:
https://knowledge.ni.com/KnowledgeArticleDetails?id=kA00Z0000019ZbJSAU
&l=sv-SE. [Accessed May 2020].
39
Appendix
40
