Recommendations for Ecg Acquisition Using Bitalino | Electrocardiography

Please download to get full document.

View again

of 6
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information Report
Category:

Documents

Published:

Views: 4 | Pages: 6

Extension: PDF | Download: 0

Share
Related documents
Description
paper on ecg acquisition for the Bitalino hardware
Transcript
  See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/308984068 RECOMMENDATIONS FOR ECG ACQUISITIONUSING BITALINO Conference Paper  · April 2016 CITATIONS 0 READS 93 5 authors , including:Lucie Mar š ánováBrno University of Technology 3   PUBLICATIONS   0   CITATIONS   SEE PROFILE Radovan SmisekBrno University of Technology 8   PUBLICATIONS   0   CITATIONS   SEE PROFILE Martin VitekBrno University of Technology 17   PUBLICATIONS   43   CITATIONS   SEE PROFILE Jana Kolá ř ováBrno University of Technology 54   PUBLICATIONS   122   CITATIONS   SEE PROFILE All content following this page was uploaded by Radovan Smisek on 11 October 2016. The user has requested enhancement of the downloaded file. All in-text references underlined in blue are added to the srcinal documentand are linked to publications on ResearchGate, letting you access and read them immediately.  RECOMMENDATIONS FOR ECG ACQUISITION USING BITALINO Andrea Němcová *, Lucie Maršánová  and Radovan Smíšek    Doctoral Degree Programme (1), *Doctoral Degree Programme (2), FEEC BUT E-mail: xnemco01@stud.feec.vutbr.cz, xmarsa08@stud.feec.vutbr.cz, xsmise00@stud.feec.vutbr.cz Supervised by: Martin Vítek    and Jana Kolářová   E-mail: vitek@feec.vutbr.cz, kolarova@feec.vutbr.cz Abstract : Cardiovascular disorders are still the most common cause of death in Western countries. In addition to cardiologists´ care, p atient self-examination is the growing area nowadays. For this  purpose, mobile devices for ECG signal acquisition are suitable. From these devices, we selected low-cost system called BITalino. In this work, we tested the quality of ECG signals acquired with BITalino under various conditions. We tested 2 values of sampling frequency, 12 electrodes  placements and 6 types of exercises. We recommend using sampling frequency of 1,000 Hz, two electrodes placements  –   on chest and on wrists for both resting and exercise ECG. Keywords : BITalino, ECG, electrocardiogram, acquisition, sampling frequency 1.   INTRODUCTION The areas of telemedicine and especially patient self-examination using mobile applications are de-veloping very fast nowadays. For this purpose, mobile and ideally wireless devices are necessary for acquiring physiological data. The data are wirelessly sent to access point (smartphone, tablet, PC), special application process them and the user is informed about their health condition. Anoth-er possibility is that the data can be sent from access point to database via the Internet and doctors evaluate them. In our work, we deal with the first part of this process, which is common for both approaches  –   data acquisition. We focus on electrocardiogram (ECG) signal only. There are several commercial systems for ECG signal acquisition (e.g. [1], [2] and [3]). For our  purpose (developing mobile application for wide spectrum of users), we need accessible, low-cost, open (we need the signal in a standard file format, such as .txt  ) and accurate system. The acquisi-tion accuracy is important because in our future work we will evaluate pulse, heart rate variability and the most common heart pathologies (premature atrial or ventricular contractions, atrial fibrilla-tions, sinoatrial blocks, atrioventricular blocks and left and right bundle branch blocks). BITalino acquisition system fulfills these conditions that is why we use it for ECG signal acquisition. More-over, the system is small, user-friendly, and on the Internet there are available APIs (Application Programming Interfaces) for different platforms for free. [3] In this work, we tested the quality of ECG signals acquired with BITalino under various conditions, namely sampling frequency, electrodes placements and quality of recorded signals. BITalino is de-scribed in chapter 2.1. The conditions are mentioned in chapters 2.2, 2.3 and 2.4. Chapter 3 deals with results. In the end, we recommend proper conditions for a good quality ECG acquisition. 2.   METHODS 2.1.   BIT ALINO ACQUISITION SYSTEM   BITalino is a multimodal system for biomedical data acquisition. It consists of microcontroller unit, Bluetooth (BT) module for wireless communication and power block with a Lithium Ion Polymer 543   battery (Figure 1). In one board, 6 various measurement sensors for bioelectrical and biomechani-cal data acquisition are integrated (Figure 1). BITalino enables acquisition of electrocardiogram, electromyogram (EMG), electrodermal activity (EDA) and accelerometry (ACC). Light sensor (LUX) and Light-Emitting Diode (LED) are included as well. Light sensor enables ambient light monitoring. Both Light sensor and LED enable synchronization with third party equipment. [4],  [5], [6] ECG signal acquisition with BITalino is noninvasive; ECG is sensed by three Ag-AgCl electrodes. BITalino enables one lead measurement with sampling frequency (f  s ) of 10/100/1,000 Hz. The  bandwidth is 0.5-40 Hz, in this range the accuracy is guaranteed. The system transfers frequencies from 40 Hz to the half of the sampling frequency as well, but with smaller magnitude. The resolu-tion of ECG signal is 10 bits. The acquired voltage range is limited to (-1.5, +1.5) mV. Data can be visualized by the software OpenSignals, which enables real-time data acquisition and offline  browsing. Recorded data can be stored in a standard ASCII file format ( .txt  ) or in the HDF5 format. [4], [5], [6]  Figure 1:   BITalino board consists of microcontroller unit (CONTROL), Bluetooth module (BT),  power block (POWER) and 6 sensors for various data acquisition. [5] 2.2.   S AMPLING FREQUENCY   BITalino generally enables to set 3 values of sampling frequency (10 Hz, 100 Hz and 1,000 Hz). For ECG acquisition, only 100 Hz and 1,000 Hz are considered. We tested both of these frequen-cies in terms of preserving the diagnostic information. 2.3.   E LECTRODES PLACEMENTS   We tested 12 electrodes placements in terms of signal quality and patient comfort. The exact loca-tion of each electrode is given in Table 1. The best electrodes placement(s) will be determined as a compromise between signal quality (amount of diagnostic information) and patient comfort. Hands are more accessible and comfortable body parts for electrodes placement than chest. On the other hand, we expected that the signal from hands has lower quality and magnitude than the signal sensed from chest. The electrodes placements were determined with respect to standard ECG leads, body part accessi- bility, assumed amplitudes and design of BITalino. The lead set is 30 cm long, which means that the distance between two electrodes (body parts) should be no more than 60 cm. From standard ECG leads, we take into consideration only limb leads (because they are bipolar). Moreover, the distance between hand and leg is greater than 60 cm, so the sensing is very uncomfortable. The on-ly standard lead, which can be used is limb lead I (left hand, right hand). The signals were acquired for 1 minute with f  s  = 1,000 Hz. The patient was sitting on chair with hands laid on thighs. All the tests in this paper were performed on three healthy subjects. 544  No. Plus Minus Reference 1 center of right palm center of left palm upper part of right palm 2 under right clavicula under left musculus pectoralis major under left clavicula 3 right wrist left wrist right wrist closer to pinky 4 right wrist left wrist right wrist closer to thumb 5 right wrist left wrist left wrist closer to pinky 6 right wrist left wrist left wrist closer to thumb 7 sternum under left pectoral muscles under sternum 8 sternum serratus anterior under sternum 9 sternum latissimus dorzi under sternum 10 sternum under left pectoral muscles serratus anterior 11 sternum latissimus dorzi serratus anterior 12 right thumb left thumb right wrist closer to pinky Table 1:   12 tested electrodes placements. 2.4.   E XERCISE ECG   We tested the signal quality while the patient was moving (exercise ECG). We tried a few types of movements, because they can influence the signal in a different way. It depends especially on elec-trodes placement. Based on chapters 2.2 and 2.3 and results 3.1 and 3.2, we set the sampling fre-quency on 1,000 Hz and choose two proper electrodes placements (No.2  –   chest and No.6  –   hands). The subject was doing each exercise for one minute, than they had 10 seconds to get ready for the next exercise. The acquisition was continual for one electrodes placement and all the exercises (to-tal 6 minutes and 50 seconds). BITalino was hanged on the neck. To quantify quality of the signal, we used segmentation algorithm [7]  based on SNR estimation. The algorithm segmented signal into 3 quality groups. The 1 st  group  –    low noise (1) has very low level of noise, which means that the signals can be delineated. In the 2 nd  group  –    medium noise (2), level of noise is higher but still enables to use the signal for QRS detection. The 3 rd  group  –    high noise (3) includes signals with small SNR; they are not appropriate for further processing. 3.   RESULTS 3.1.   T ESTING OF SAMPLING FREQUENCY   The acquired signals were compared visually by two experts in terms of details. In case of f  s  = 1,000 Hz, more details are preserved, but the signal is 10 ×  longer. In case of f  s  = 100 Hz the signal is smoother, but some details are lost (rapid changes). These details are important for diag-nosing some pathologies (e.g. bundle branch blocks). The sampling frequency should be set with respect to required diagnostic information. F s  = 100 Hz is enough for rhythm monitoring. For mor- phology assessment, f  s  = 1,000 Hz is necessary. 3.2.   T ESTING OF ELECTRODES PLACEMENTS   After signals acquisition, we evaluated them manually (results are shown in Table 2) in terms of:     presence of three important components (P, R, T)  –   0 means absence, 1 means presence and the component is quite small, 2 means presence and the component is significant (eve-ry component in each cycle always belongs in one group (0, 1, 2) within one signal),    trimming of R wave  –   Y (yes)/N (no),    amplitudes of R and T,    noise - 0 absence, 1 very low level, 2 low level, 3 high level (different scale than in the segmentation algorithm [7]),    isoline  –   shift of isoline, drift (D). 545
Recommended
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks