Available olie www.jsaer.com, 2018, 5(7):1-7 Research Article ISSN: 2394-2630 CODEN(USA): JSERBR 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 Compariso of Differet Cotact Material o ECG Sigal Nik Tehrai Iteratioal Techological Uiversity, Sa Jose, Califoria Abstract This test is desiged to characterize ad compare the sigal quality of traditioal wet (gel) electrode cotact material vs. the dry (stailess steel) electrode cotact material used i the device after post processig of collected raw data from each electrode type. Two types of cotact electrode materials were used: 1) Electrode Pair A- Hydrogel Electrodes: MEDI-TRACE(R) 500 ECG ELECTRODE (K945479), ad 2) Electrode Pair B- Stailess Steel Electrodes: 136 SS of fial device.after froted filterig ad baselie removal, both pairs (A&B) are qualitatively similar. This is maily due to the fact that SS teds to have more differetial baselie movemet caused by higher cotact resistace to the ski. Results show that after sigal processig (icludig badpass filterig ad additioal baselie removal), two simultaeously captured ECG sigals, the first oe usig HG electrodes, ad the secod usig SS electrodes, are equivalet i terms of SNR ad correlatio. Although both this device (K172654) ad predicate (K150869) use as dry electrodes ad are equivalet, this report provides additioal compariso of wet (gel) electrode with dry (stailess steel). Keywords Sigal quality wet (gel) vs. dry (stailess steel) electrodes, ECG sigals, HG electrodes. Itroductio The purpose of this test coducted by DyoSese Corp. was to characterize ad compare the sigal quality of traditioal wet (gel) electrode cotact material vs. the dry (stailess steel) electrode cotact material used i the device after post processig of collected raw data from each electrode type. Test Setup ad Data Collectio Two differet types of cotact electrode materials were used; Electrode Pair A- Hydrogel Electrodes: MEDI-TRACE(R) 500 ECG ELECTRODE (K945479) Electrode Pair B- Stailess Steel Electrodes: 136 SS of fial device Data collectio was carried out for both cotact types simultaeously usig idetical device hardware versio of the fial product (hardware: 01051-02, 01051-03, Firmware: 0.3.51 Software: 2.0.21) Each electrode pair A ad B were placed at same locatio ext to each other durig simultaeous capture. The device was coected to a PC with a Bluetooth dogle to capture ad store the data. Each capture data is the post processed i exact techique used i fial device (see below). Capture ad Post Processig Descriptio For each electrode type, the ECG sigal is captured through same itegrated aalog froted circuit (IC) of the device that cosists of three istrumetatio amplifiers to reduce large direct curret (DC) chages ad widebad oise. The sigal is digitized with a 24-bit aalog-to-digital coversio (ADC) at a samplig rate of 500 Hz ad stored as raw sigal. This raw data is badpass filtered i the rage of 0.05Hz-100Hz followed by additioal baselie removal algorithm. 1
Tehrai N, 2018, 5(7):1-7 Figure 1: Example of captured raw sigal for each electrode pair A ad B. Figure 2: Example after badpass filtered of each electrode pair A ad B. 40 41 42 43 Figure 3: Example after complete processig for additioal baselie removal for each electrode pair A ad B. As show i Figure 3, after froted filterig ad baselie removal, both pairs (A&B) are qualitatively similar. This is maily due to the fact that SS teds to have more differetial baselie movemet caused by higher cotact resistace to the ski, as show i Figure 1. 2
Tehrai N, 2018, 5(7):1-7 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 Populatio Study Descriptio Twety oe voluteers participated i the study with matchig the U.S.demographic distributioprovided i table 1 below [1-2]. Table 1: Demographic Number of subjects 21 Age 19-25 2 26-54 11 55+ 8 Sex Male 11 Female 10 Ethicity White 14 No-white 7 (Africa America, Hispaic, Asia, mixed Race, etc.) Simultaeous ECG measuremets were captured usig SS ad Stadard Gel electrode from all twety oe (21) subjects. Sigal Quality Aalysis To study if the sigal quality of dry SS electrode (Pair B) is comparable to the stadard wet electrodes (Pair A), the SNR ad correlatio betwee these two electrodes are compared for the collected populatio. Sigal-to-Noise (SNR) Compariso Sigal (QRS template) defiitio: The ECG sigal (oise free) is defied by a QRS template derived from the media (average) of idividual QRS of the processed ECG (Figure 4, show as a solid blue color). Noise defiitio: The oise is defied by the differece betwee the QRS template ad each idividual QRS i etire ECG sigal (Figure 4, defied as Noise = (A B)) Sigal-to-oise defiitio: The SNR is the ratio of the power of the sigal (QRS template) to power of the oise. 63 64 65 66 Figure 4: The clea sigal is defied by the media template beat. The oise is defied by the differece betwee idividual beats ad the template 3
Tehrai N, 2018, 5(7):1-7 67 68 69 70 71 72 73 74 Table 2 (ad Figure 5), below, shows computed SNR for each subject for Pair A ad Pair Bad the differeces (Pair A-Pair B). We also computed the mea ad stadard deviatio across all populatio. Table 2: SNR ratios ad mea calculatios for hydrogel ad stailess steel for all 21 records (see Figure 5) Subjects Pair A Pair B Pair A- Pair B Hydrogel electrodes Stailess Steel SNR Differece SNR S1 23.1 23.6-0.5 S2 2.9 3.3-0.5 S3 9.8 6.56 3.3 S4 6 6.9-0.9 S5 11.8 10.8 1.0 S6 2.6 1.3 1.3 S7 6.4 3.3 3.1 S8 4.6 6.9-2.4 S9 16.3 12.9 3.4 S10 25.4 26-0.6 S11 22.8 21.1 1.7 S12 15.1 13.8 1.2 S13 13.9 9.2 4.6 S14 7 6.5 0.5 S15 2.2 2.2 0.0 S16 9.6 6.4 3.2 S17 3.3 3.6-0.3 S18 15.2 12.8 2.4 S19 9.5 6.6 2.9 S20 9.4 7.7 1.7 S21 5.4 5.1 0.3 Mea 10.6 9.4 1.2 Std 7 6.9 1.8 Edpoits ad success criteria for SNR result Cofidece iterval for the true mea differece was computed to defie the accuracy of the estimate ad establish acceptable limits for the mea differeces (true mea differece). Cofidece iterval for the true mea differece is calculated as below: d ± t Sd 75 76 77 78 79 d is the mea vector of differeces betwee hydrogel ad stailess steel measures. Sd is the stadard deviatio of the differeces Sd is the stadard error of the mea differece Summary Aalysis for Differece d Mea differece Sd stadard deviatio of the differeces Sd stadard error of the mea differece mea 21 1.2 1.8 0.393 With the acceptable assumptio of 95% cofidece ad o ( 1) degrees of freedom, cofidece iterval for the true mea differece is therefore: Cofidece Iterval = d ± 2.086 Sd = = 0.381, 2.019 4
Tehrai N, 2018, 5(7):1-7 80 81 82 83 84 85 86 This cofirms that with 95% level of cofidece, we ca be sure the mea differece i SNR betwee Pair A ad B is lower tha differece across populatio of each Pair A or B ad lies somewhere betwee just 0.381 ad 2.019 ad summarized below. 95% Cofidece Iterval Degree of Mea t lower upper Differece freedom (df) differece 20 2.086 1.2 0.381 2.019 Note that the SNR degradatio i rage of 0.381 to 2.019is smaller compared to the stadard deviatio of the measures of either Pair A (std: 7) or Pair B (std: 6.9), provig that differece betwee cotact materials has o meaigful SNR impact whe compared agaist across all populatios. Pair A: Blue lie Pair B: Red lie 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 Figure 3: Liear SNR Correlatio Coefficiet Compariso We assess the similarity of HG electrodes ad SS electrodes ECG captures by computig the Pearso correlatio coefficiet betwee both Pair A ad Pair B: Pearso correlatio coefficiet Equatio N i=1 x i x y i y r =, N x i x 2 i=1 N i=1 y i y 2 N is the sample size x i, y i are the sigle samples idexed with i x = 1 N N i=1 x i, (the sample mea); ad aalogously fory. For each subject a iterval of 60 secs for a data samplig rate of 500Hz is used to calculate correlatio coefficiets which correspods to 30000 data poits (N=Sample size i Pearso correlatio coefficiet formula). The correlatio coefficiet was computed usig Pearso correlatio coefficiet for each subject usig both electrodes. The results are listed i Table 3: 5
Tehrai N, 2018, 5(7):1-7 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 Table 3: Correlatio betwee ECG records from wet get (Pair A) ad stailess steel (Pair B) for all 21ECG records Subject Correlatio S1 0.972 S2 0.884 S3 0.938 S4 0.951 S5 0.959 S6 0.788 S7 0.897 S8 0.922 S9 0.963 S10 0.975 S11 0.978 S12 0.969 S13 0.952 S14 0.893 S15 0.864 S16 0.925 S17 0.897 S18 0.948 S19 0.9298 S20 0.9405 S21 0.9254 Mea correlatio 0.927 Std of the correlatio 0.045 Sigals of both pair of electrodes show mea correlatio of 0.927cofirmig a strog correlatio betwee both types of ECG electrodes (correlatio above 0.7 is reported as threshold of strog correlatio) [1]. 1 Additioally the subject havig the highest oise (S6) resulted i a correlatio of 0.788, which still meets the acceptace criteria. Coclusios Results show that after sigal processig (icludig badpass filterig ad additioal baselie removal), two simultaeously captured ECG sigals, the first oe usig HG electrodes, ad the secod usig SS electrodes, are equivalet i terms of SNR ad correlatio. Although both this device (K172654) ad predicate (K150869) use as dry electrodes ad thus are equivalet, this report provides additioal compariso of wet (gel) electrode with dry (stailess steel) to further show use of dry electrode does ot adversely affect performace of captured ECG waveform. Refereces [1]. Ecyclopedia Britaica. (2018). Populatio: Age Distributio. Biology ad athropology. Retrieved from https://www.britaica.com/sciece/populatio-biology-ad-athropology/age-distributio. [2]. Ecyclopedia Britaica. (2018). Ethic or racial compositio. Biology ad athropology. Retrieved from https://www.britaica.com/sciece/populatio-biology-ad-athropology/ethic or racial compositio [3]. Che, Y.H., de Beeck, M., Vaderheyde, L., Carrette, E., Mihajlović, V., Vastreels, K.Grudleher, B., et al. (2014). Soft, Comfortable Polymer Dry Electrodes for High Quality ECG ad EEG Recordig. Sesors, 14(12), 23758 23780. MDPI AG. Retrieved from http://dx.doi.org/ 10.3390/s141223758 6