Available online at www.sciencedirect.com
ScienceDirect Journal of Electrocardiology xx (2014) xxx – xxx www.jecgonline.com
Surface 12 lead electrocardiogram recordings using smart phone technology Giselle A. Baquero, MD, a Javier E. Banchs, MD, b Shameer Ahmed, MD, a Gerald V. Naccarelli, MD, a Jerry C. Luck, MD a,⁎ a
Penn State Hershey Heart & Vascular Institute, Division of Cardiology, Department of Medicine, Penn State College of Medicine, Penn State University, Hershey, PA b Texas A & M College of Medicine, Baylor Scott & White Health, Temple, TX
Abstract
Importance: AliveCor ECG is an FDA approved ambulatory cardiac rhythm monitor that records a single channel (lead I) ECG rhythm strip using an iPhone. In the past few years, the use of smartphones and tablets with health related applications has significantly proliferated. Objective: In this initial feasibility trial, we attempted to reproduce the 12 lead ECG using the bipolar arrangement of the AliveCor monitor coupled to smart phone technology. Methods: We used the AliveCor heart monitor coupled with an iPhone cellular phone and the AliveECG application (APP) in 5 individuals. Results: In our 5 individuals, recordings from both a standard 12 lead ECG and the AliveCor generated 12 lead ECG had the same interpretation. Conclusions: This study demonstrates the feasibility of creating a 12 lead ECG with a smart phone. The validity of the recordings would seem to suggest that this technology could become an important useful tool for clinical use. This new hand held smart phone 12 lead ECG recorder needs further development and validation. © 2014 Elsevier Inc. All rights reserved.
Keywords:
AliveCor ECG; Electrocardiogram; 12 lead; Smart phone ECG
Introduction The standard 12 lead electrocardiogram (ECG) has been used clinically for over 100 years for the diagnosis of ischemic heart disease and cardiac arrhythmias [1–3]. New technology today allows for a standardized single lead recording to be obtained by a computerized monitoring bipolar device that is coupled and wirelessly connected to a smart phone [4–7]. The standard bipolar limb leads of the 12 lead ECG can be theoretically reproduced with simple bipolar connections of the limbs according to Einthoven's triangle [1,2]. The augmented unipolar limb leads (VR, VL, VF) can also be re-created theoretically by connecting the exploring electrode to the focus limb and the other two limbs to the negative pole of the recording bipolar device [3]. Similarly, the unipolar chest leads can be created with a bipolar arrangement: the positive roving electrode is on the chest (at V1, V2, V3, V4, V5, V6) and the distant negative electrode is on the lower left leg, some distance from the heart. In this initial feasibility trial, we attempted to ⁎ Corresponding author at: 500 University Drive, P.O. Box 850. H047, Hershey, PA 17033–0850. E-mail address: [email protected] http://dx.doi.org/10.1016/j.jelectrocard.2014.09.006 0022-0736/© 2014 Elsevier Inc. All rights reserved.
reproduce the 12 lead ECG using the bipolar arrangement of the AliveCor monitor coupled to smart phone technology.
Methods Patient population We used the AliveCor heart monitor coupled with an iPhone cellular phone and the AliveECG application (APP). The population for this smart phone 12-lead ECG feasibility study included 5 individuals: two normal individuals with a narrow QRS complex and three with a wide QRS complex: one each with right bundle branch block, left bundle branch block, and dextrocardia (Table 1). Each of these individuals had a standard 12 lead ECG performed with a MAC5500, multichannel ECG machine. In case #1 with a narrow QRS and case #2 with a left bundle branch block, the standard 12 lead ECG and the smart phone created ECG were recorded at the same sitting and using the same electrode placements on the arms, left lower leg just above the ankle and the precordial leads. Case #3 is a healthy competitive cyclist who had the smart phone generated 12 lead ECG performed
2
G.A. Baquero et al. / Journal of Electrocardiology xx (2014) xxx–xxx
by the authors and one week later had a standardized ECG performed by a heart station technician. Case #4 had the standard tracing performed one week prior to surgical removal of a esophageal cancer and the smart phone tracing performed on day 4 post operation. Case #5 has dextrocardia and had the standard tracing on admission for endocarditis and the smart phone tracing performed 2 days later.
LEADS I, II, III RIGHT
LEFT
-
+-
-
+
AliveCor technology and modification The AliveCor device consist of FDA approved (1) hardware, a bipolar electrode case that fits on a smart phone to record the hearts electrical activity and (2) software to process, save and transmit the single lead electrocardiogram. An APP on the smart phone (AliveECG) transforms the electrical signal to ultrasound. The smart phone's microphone will receive the ultrasound signal and transmit the electrocardiogram as a wireless PDF format.
+
+
LEADS aVR, aVL, aVF
Smart phone ECG recording RIGHT
The AliveCor device is a standardized, bipolar single lead recorder. Using the principles of Einthoven's triangle, the standard ECG bipolar and unipolar limb leads can be reproduced. To record the standard bipolar limb leads; the AliveCor case was adapted with adhesive ECG tabs (Kendall, MT710 Diagnostic Tab Electrodes). These standard adhesive tabs are placed on the right and left forearms and the lower left leg just above the ankle. The adhesive tabs are used for the precordial leads (V1, V2, V3, V4, V5, V6) and are placed on the chest using the standard locations [1,2]. Using three 24 inch insulated wires with mini alligator clips, the standard bipolar (leads: I, II, III) and unipolar (aVR, aVL, aVF) limb leads and the precordial leads are recorded in successive fashion (Fig. 1). For the augmented unipolar limb leads, we used a positive lead as the exploring lead and the other two limb leads are connected to the negative pole. For the six chest leads, the positive electrode is the exploring electrode while the indifferent lead is negative and is located on the distant left leg. Using the principles of the Einthoven's triangle and Goldberger modification, the standardized AliveCor ECG can be modified to obtain a 12 lead ECG recording [2,3]. A standard 12 lead recording was made using a Mac 5500 recorder. In case #1 and case #2, the authors performed the standard 12 lead recording and used the same exact sites for the smart phone recording. In case #3, case #4 and case #5, the standard Mac 5500 12 lead ECG was performed by an ECG technician. The precordial lead electrode placements were not exactly the same sites as those used for the AliveCor recordings that were done at a different setting. Results of the 5 individual recordings The clinical characteristics of the 5 individuals in this feasibility study are described in Table 1. Case #1 is a 67 year old male with a normal echocardiogram, ejection fraction 0.55, normal chamber sizes, and normal valve function. A standard 12 lead electrocardiogram (ECG) and then a sequential AliveCor 12 lead ECG are performed using the same lead tabs on the limbs and chest.
LEFT
+-
+-
-
+
-+
LEADS V1-V6 RIGHT
LEFT V1 - V6
++ ++++
-
+
Fig. 1. AliveCor ECG recording. Panel A: Bipolar limb lead recordings (I, II, III) using Einthoven's triangle format. Panel B: Unipolar limb lead recordings (aVR, aVL, aVF) using the method of Goldberger with the focus exploring electrode as positive and the other two limbs as the negative. Panel C: Precordial leads V1–V6: with the positive exploring electrode on the chest and the negative electrode on the distant lower left leg.
Both the Mac 5500 standard 12 lead ECG and the AliveCor generated 12 lead ECG have the same interpretation. Both show sinus rhythm, normal intervals; normal axis + 60 degrees; left atrial abnormality and left ventricular hypertrophy by voltage. The AliveCor tracing is a sequential obtained ECG and demonstrates the presence of a sinus arrhythmia. In Fig. 2, there are comparisons of the two tracings. The email PDF transfer of the AliveCor version is reconstructed for printing the 12 lead. Despite the
G.A. Baquero et al. / Journal of Electrocardiology xx (2014) xxx–xxx Table 1 Clinical characteristics of the study individuals. Case
Age/sex
Clinical diagnosis
Baseline ECG
#1 #2 #3 #4 #5
67/M 55/M 23/M 70/M 31/M
APC’S, VPC’S Cardiomyopathy none Esophageal CA Endocarditis
Sinus rhythm rate 60; LVH, LAA Sinus rhythm rate 96; LBBB Sinus rhythm rate 70; normal Sinus tachycardia rate 116; RBBB Atrial fibrillation; Dextrocardia Junctional rhythm rate 71
APC’S = atrial premature complexes. VPC’S = ventricular premature complexes. LAA = left atrial abnormality. LVH = left ventricular hypertrophy. LBBB = left bundle branch block. RBBB = right bundle branch block.
reconstruction, the QRS amplitudes, axis, intervals, initial and terminal QRS forces and “P” wave and “T” wave vectors are remarkably similar. Case #2 is a 55 year old male with a tachycardia-induced cardiomyopathy secondary to atrial fibrillation with a rapid ventricular rate. He is now post direct current cardioversion and has symptomatic improvement in sinus rhythm. He has an underlying left bundle branch block. A standard 12 lead ECG and then the sequential AliveCor 12 lead ECG are performed using the same lead tabs on the limbs and chest. Both ECG tracings show sinus rhythm, normal PR and QT intervals but a wide left bundle branch block QRS complex and left atrial abnormality. The standard ECG and the
3
sequentially obtained AliveCor generated 12 lead ECG are compared in Fig. 3. The QRS amplitudes, axis, intervals, initial and terminal QRS forces and “P” wave and “T” wave vectors mimic the standard ECG recording. In the smart phone recording, precordial leads V1, V2 and V3 show higher ST elevation than in the Mac 5500 recording. Case #3 is a 23 year old male and is a competitive cyclist with a normal echocardiogram, ejection fraction 65%. A standard 12 lead ECG and the AliveCor generated ECG are performed on two different days. The chest leads position V1–V4 are the same but V5 and V6 are different. Both the standard ECG and the AliveCor ECG have the same interpretation. They show sinus rhythm, normal intervals, and a vertical axis. Both are normal ECGs. In Fig. 4, both tracings are compared. The limb leads QRS complexes amplitudes, axis, intervals and initial and terminal forces mimic each other on the standard and AliveCor tracings. The vector of the “P” wave is the same between the two. The chest leads V1, V2, V3 and V4 are quite similar but V5 and V6 are different given different positions. In the smart phone recording, precordial leads V1, V2, V3 show J point elevation and ST elevation that is not seen in the Mac 5500 recording. Case #4 is a 70 year old symptomatic male with shortness of breath. There is a prior history of coronary artery disease, right bundle branch block and now esophageal carcinoma. The standard 12 lead ECG is a pre-operation recording and
Fig. 2. Comparison of the standard and device generated 12 lead recordings for case #1: Top trace Mac 5500 12 lead ECG and below AliveCor ECG with normal standardization at 25 mm/second (S): Interpretation standard: sinus rhythm, rate 60 beats/minute (bpm), normal intervals: PR 0.17 s, QRS duration 0.08 s, QT 0.39 s, QRS axis +60 degrees. Left atrial abnormality, voltage criteria for left ventricular hypertrophy. AliveCor generated ECG from sequentially transmitted PDF files and then assembled as shown: Interpretation: sinus rhythm and sinus arrhythmia, rate varies 56–71 bpm, normal conduction intervals: PR 0.17 s, QRS duration 0.08 s, QT 0.4 s, QRS axis +55 degrees. Left atrial abnormality, voltage criteria for left ventricular hypertrophy. A comparison shows a true 12 for 12 ECG lead match in these ECGs recorded nearly simultaneously from the same electrode tabs. A comparison of the QRS, “P” wave and “T” wave amplitudes and axis is the same between the two ECGs.
4
G.A. Baquero et al. / Journal of Electrocardiology xx (2014) xxx–xxx
Fig. 3. Comparison of the standard and device generated 12 lead recordings for case #2: Top trace standard ECG and below device generated ECG: Interpretation standard: sinus rhythm rate 94 bpm; PR 0.16 s, QRS duration 0.17 s, QT 0.38 s, QRS axis −15 degrees; left atrial abnormality; left bundle branch block with repolarization changes. Device generated ECG from sequentially transmitted PDF files and reassembled as shown: Interpretation: sinus rhythm, rate 100 bpm; PR 0.16 s; QRS duration 0.16 s; QT 0.4 s; QRS axis −15 degrees. Left atrial abnormality; left bundle branch block with repolarization changes. The comparison shows a true 11 for 12 ECG lead match in these ECGs recorded sequentially using the same limb and precordial position electrode tabs. Only lead aVF is slightly different.
Fig. 4. Comparison of the standard and device generated 12 lead recordings for case #3: (same format as above): Interpretation standard: sinus rhythm and sinus arrhythmia rate varies from 62 bpm to 88 bpm, PR 0.12 s, QRS duration 0.11 s, QT 0.4 s, QRS axis +100 degrees. Normal ECG. Device generated ECG interpretation: sinus rhythm rate 48 bpm, PR 0.12 s, QRS duration 0.10 s, QT 0.4 s, QRS axis +100 degrees. Normal ECG. In this comparison, the limb leads from the two formats mimic each other. The chest leads, V1, V2, V3 and V4 are the same but leads V5 and V6 are different given the lead placements are different.
G.A. Baquero et al. / Journal of Electrocardiology xx (2014) xxx–xxx
5
Fig. 5. Comparison of the standard and device generated 12 lead recordings for case #4: (same format as above): Interpretation standard: sinus tachycardia, rate 116 bpm, PR 0.15 s, QRS duration 0.16 s, QT 0.38 s, QRS axis − 45 degrees. Left axis deviation; right bundle branch block with repolarization abnormality; inferior myocardial infarction, age undetermined. Device generated ECG interpretation: sinus rhythm rate 75 bpm, PR 0.18 s, QRS duration 0.13 s, QT 0.4 s, QRS axis − 45 degrees. Left axis deviation, right bundle branch block with repolarization abnormality, inferior myocardial infarction. Despite the different chest lead positions and the different days for each, the two are quite similar in most leads.
the AliveCor generated 12 lead is a postoperative recording with different chest positions. The standard 12 lead ECG shows sinus tachycardia, rate 116 bpm; wide QRS duration, normal QT, left axis deviation, − 45 degrees, right bundle branch block with a repolarization abnormality and an inferior myocardial infarction, age undetermined (Fig. 5). The AliveCor generated 12 lead ECG also shows sinus rhythm, left axis deviation, right bundle branch block and the age undetermined inferior infarct. Although the tracings are from different days, there is a remarkable similarity between the two recordings QRS durations, amplitudes and axis. The initial and terminal QRS forces and “P” wave vectors are also similar. There are ST depressions that seem exaggerated in V1 and V2 in the smart phone recording. Case #5 is a 31 year old male symptomatic with fever, malaise, shortness of breath, and a new diagnosis of endocarditis. He has a long-standing permanent pacemaker with a trans-venous right atrial pacing electrode and an epicardial ventricular pacing lead. The base line cardiac diagnoses are persistent atrial fibrillation, complete atrialventricular block and congenitally corrected transposition of the great arteries with dextrocardia. The standard 12 lead ECG and the AliveCor generated 12 lead are from different days. The standard 12 lead ECG shows underlying atrial fibrillation, an accelerated junctional rhythm at 72 bpm and dextrocardia with right axis deviation and poor R wave progression. In Fig. 6, the AliveCor generated 12 lead ECG shows atrial fibrillation, an accelerated junctional rhythm
and a comparable QRS morphology that also supports the diagnosis of dextrocardia. In the smart phone recording, there are repolarization differences in leads I, II, V1, V2 and V4–V6.
Discussion The standard 12-lead electrocardiogram remains the backbone for clinical decision-making in the acute care arena particularly for diagnosis and treatment of the acute coronary syndromes. It remains the primary tool for the recognition of and the diagnosis of cardiac arrhythmia. In both cases, the earlier the 12 lead ECG can be obtained, the earlier the diagnosis, initiation of therapy and the reduction in morbidity and mortality. The new technology is moving medicine in that direction. The AliveCor single channel validated and standardized ECG recorder for smart phone use is a step in the direction of better technology. We hypothesize from our 5 case series that the device can be easily modified to become a small, versatile and practical 12 lead ECG recorder. The smart phone is a ubiquitous part of life today and for it to have the potential to easily record a simple important medical test such as the 12 lead ECG may be potentially lifesaving. In this report, we show the feasibility of converting a smart phone based bipolar heart rhythm recorder to a 12 lead recorder with standard ECG tabs and wires with alligator
6
G.A. Baquero et al. / Journal of Electrocardiology xx (2014) xxx–xxx
Fig. 6. Comparison of the standard and device generated 12 lead recordings for case #5: (same format as above): Interpretation standard: underlying atrial fibrillation, a regular wide QRS with a QRS duration 0.12 s (non-specific intra-ventricular conduction delay), QT 0.38 s, QRS axis +150 degrees. An accelerated junctional rhythm at 72 bpm. There is right axis deviation, poor R wave progression and reduced QRS voltages in the left lateral precordial leads that are consistent with dextrocardia. Device generated ECG interpretation: Underlying atrial fibrillation, regular wide QRS rhythm, QRS duration 0.12 s, QT 0.44 s, QRS axis +150 degrees. Accelerated junctional rhythm, at 64 bpm, right axis deviation, poor R wave progression without transition consistent with dextrocardia. The ECGs are quite comparable despite different precordial lead placement and on different days.
clips from a hardware store. We simply followed the tenets of Einthoven's triangle to sequentially perform a standardized 12 lead ECG in 5 individuals. The bipolar limb lead tracings (I, II, III) are easily recorded with the AliveCor hardware. Using the semi-direct lead hypothesis of electrocardiography, we could use an “exploring electrode” on the chest near the heart's surface and a distant “indifferent electrode” on the lower left leg to record the unipolar chest leads (V1, V2, V3, V4, V5, V6) [1]. In the absence of the “Wilson central terminal,” we recorded the unipolar limb leads (aVR, aVL, aVF) with an exploring electrode on the focused limb and the other two limbs as the composite indifferent electrode. This is the Goldberger unipolar limb lead arrangement [3]. We show five examples of 12 lead ECGs recorded with the use of a smart phone and the AliveCor technology. In its present form, the AliveCor technology is different from the Mac 5500 technology. The most obvious is the ease at which the ECG is obtained on a 3-channel recorder and printed in a standard 12 lead format. The AliveCor recorder is a single channel, bipolar platform that we adapted to construct a 12 lead tracing. The AliveCor does not have a “Wilson central terminal” which is a standard feature of the Mac 5500 recorder. This feature allows for effectively separating a bipolar signal into two components and allows for the unipolar recordings. The central terminal stabilizes the
electrical forces at the indifferent electrode by creating a “zero potential”. This allows the unipolar exploring electrode to accurately depict the focal electrical activity without the interference of even a distant indifferent lead [1,2]. The AliveCor presently does not have this feature. This probably explains the differences in the magnitude of some of the repolarization differences in the smart phone recordings. These differences are magnified in the unipolar precordial leads. This is seen in case #2 (simultaneous recordings) where there is ST elevation in the precordial leads that is absent in the Mac 5500 recording. Most differences in repolarization (i.e. ST-T abnormalities) reflect the fact that in cases #3, #4 and #5, the ECG recordings were done on different days and the precordial leads are placed differently. When it comes to depolarization (QRS morphology), the accuracy of the smart phone recording appears to mimic the standard tracings in all aspects. This is particularly so in the two individuals (cases #1 and #2) where the recordings were sequential using the same leads. In the other three individuals the limb leads mimic the standard recordings but the chest leads have some differences possibly due to differing chest lead positions at the time of the two recordings on different days. This series includes three individuals with abnormal QRS complex morphology: LBBB, RBBB and dextrocardia. All three are accurately depicted by the smart phone version of the ECG. Depolarization is a shorter and less complex
G.A. Baquero et al. / Journal of Electrocardiology xx (2014) xxx–xxx
electrical event than repolarization [1]. Despite the lack of a “Wilson central terminal”, depolarization seems to be accurately replicated by the smart phone. Repolarization is prolonged and complex and it is easily influenced by any change in positional, hemodynamic, metabolic and physiological factors [1]. The populations that may find this “mobile 12 lead ECG” useful might include those patients with a prior myocardial infarction and those with documented arrhythmias. Patients with known coronary disease may develop a clinical event remote from a hospital but could record and transmit a 12 lead ECG. Those with known arrhythmias who develop symptoms can record and transmit not just the single lead event monitor strip but also a 12 lead of the tachycardia for diagnostic review. The populations that may benefit the most are in the third world where resources are limited but a smart phone is available. A smart phone 12 lead ECG in remote 3rd world locations could revolutionize the recognition of acute cardiac issues in those without access to healthcare in a more traditional setting. Despite the obvious limitations of different precordial lead positions in some individuals, the study demonstrates the feasibility of creating a 12 lead ECG with your smart phone. A minor limitation is the time it takes to construct the 12 lead recording on the smart phone. It takes us about 10– 12 minutes to obtain the 12 single channel strips and place them into 12 PDF files. This time restraint may be reduced by 50% once the learning curve has been eliminated. The tracings are easily visualized on the smart phone's screen separately. They can be easily transmitted as separate PDF files for review on another device and then printed in separate strips. The quality of the recordings would seem to
7
suggest that this technology could become an important useful tool for clinical use. The series of 5 cases is small but each supports the concept that the smart phone technology with its virtual 12 lead ECG can reliably recreate a reliable QRS complex. This does not validate the concept but merely presents it as plausible. Design of a new smart phone case (hard ware upgrade) seems a small challenge. The APPs and software upgrades to make the leads acquisitions simultaneous and easily transferable seem very plausible. The addition of a “Wilson central terminal” is also plausible and may help with stabilization of repolarization on the unipolar recordings. This new hand held smart phone 12 lead ECG recorder needs further development and validation. References [1] Lipman BS, Massie E, Kleiger RE. Clinical scalar electrocardiography. Chicago: Year Book Medical Publishers, Inc.; 197322–31. [2] Marriott HJL. Practical electrocardiography. Baltimore/London: Williams and Wilkins; 19801–7. [3] Goldberger E. Simple electrocardiographic electrode of zero potential and technic of obtaining augmented unipolar extremity leads. Am Heart J 1942;23:483–92. [4] Saxon LA, Smith A, Doshi S, Albert D. I phone rhythm strip. The implications of wireless and ubiquitous heart rate monitoring. J Am Coll Cardiol 2012;59(13s1):E726-E726. [5] Saxon LA. Ubiquitous wireless ECG recording: powerful tool physicians should embrace. J Cardiovasc Electrophysiol 2013;24 (4):480–3. [6] Mittal S, Movsowitz C, Steinberg JS. Ambulatory external electrocardiographic monitoring: focus on atrial fibrillation. J Am Coll Cardiol 2011;58(17):1741–9. [7] Kamel H, Smith WS. Detection of atrial fibrillation and secondary stroke prevention using telemetry and ambulatory cardiac monitoring. Curr Atheroscler Rep 2011;13(4):338–43.
ScienceDirect Journal of Electrocardiology xx (2014) xxx – xxx www.jecgonline.com
Surface 12 lead electrocardiogram recordings using smart phone technology Giselle A. Baquero, MD, a Javier E. Banchs, MD, b Shameer Ahmed, MD, a Gerald V. Naccarelli, MD, a Jerry C. Luck, MD a,⁎ a
Penn State Hershey Heart & Vascular Institute, Division of Cardiology, Department of Medicine, Penn State College of Medicine, Penn State University, Hershey, PA b Texas A & M College of Medicine, Baylor Scott & White Health, Temple, TX
Abstract
Importance: AliveCor ECG is an FDA approved ambulatory cardiac rhythm monitor that records a single channel (lead I) ECG rhythm strip using an iPhone. In the past few years, the use of smartphones and tablets with health related applications has significantly proliferated. Objective: In this initial feasibility trial, we attempted to reproduce the 12 lead ECG using the bipolar arrangement of the AliveCor monitor coupled to smart phone technology. Methods: We used the AliveCor heart monitor coupled with an iPhone cellular phone and the AliveECG application (APP) in 5 individuals. Results: In our 5 individuals, recordings from both a standard 12 lead ECG and the AliveCor generated 12 lead ECG had the same interpretation. Conclusions: This study demonstrates the feasibility of creating a 12 lead ECG with a smart phone. The validity of the recordings would seem to suggest that this technology could become an important useful tool for clinical use. This new hand held smart phone 12 lead ECG recorder needs further development and validation. © 2014 Elsevier Inc. All rights reserved.
Keywords:
AliveCor ECG; Electrocardiogram; 12 lead; Smart phone ECG
Introduction The standard 12 lead electrocardiogram (ECG) has been used clinically for over 100 years for the diagnosis of ischemic heart disease and cardiac arrhythmias [1–3]. New technology today allows for a standardized single lead recording to be obtained by a computerized monitoring bipolar device that is coupled and wirelessly connected to a smart phone [4–7]. The standard bipolar limb leads of the 12 lead ECG can be theoretically reproduced with simple bipolar connections of the limbs according to Einthoven's triangle [1,2]. The augmented unipolar limb leads (VR, VL, VF) can also be re-created theoretically by connecting the exploring electrode to the focus limb and the other two limbs to the negative pole of the recording bipolar device [3]. Similarly, the unipolar chest leads can be created with a bipolar arrangement: the positive roving electrode is on the chest (at V1, V2, V3, V4, V5, V6) and the distant negative electrode is on the lower left leg, some distance from the heart. In this initial feasibility trial, we attempted to ⁎ Corresponding author at: 500 University Drive, P.O. Box 850. H047, Hershey, PA 17033–0850. E-mail address: [email protected] http://dx.doi.org/10.1016/j.jelectrocard.2014.09.006 0022-0736/© 2014 Elsevier Inc. All rights reserved.
reproduce the 12 lead ECG using the bipolar arrangement of the AliveCor monitor coupled to smart phone technology.
Methods Patient population We used the AliveCor heart monitor coupled with an iPhone cellular phone and the AliveECG application (APP). The population for this smart phone 12-lead ECG feasibility study included 5 individuals: two normal individuals with a narrow QRS complex and three with a wide QRS complex: one each with right bundle branch block, left bundle branch block, and dextrocardia (Table 1). Each of these individuals had a standard 12 lead ECG performed with a MAC5500, multichannel ECG machine. In case #1 with a narrow QRS and case #2 with a left bundle branch block, the standard 12 lead ECG and the smart phone created ECG were recorded at the same sitting and using the same electrode placements on the arms, left lower leg just above the ankle and the precordial leads. Case #3 is a healthy competitive cyclist who had the smart phone generated 12 lead ECG performed
2
G.A. Baquero et al. / Journal of Electrocardiology xx (2014) xxx–xxx
by the authors and one week later had a standardized ECG performed by a heart station technician. Case #4 had the standard tracing performed one week prior to surgical removal of a esophageal cancer and the smart phone tracing performed on day 4 post operation. Case #5 has dextrocardia and had the standard tracing on admission for endocarditis and the smart phone tracing performed 2 days later.
LEADS I, II, III RIGHT
LEFT
-
+-
-
+
AliveCor technology and modification The AliveCor device consist of FDA approved (1) hardware, a bipolar electrode case that fits on a smart phone to record the hearts electrical activity and (2) software to process, save and transmit the single lead electrocardiogram. An APP on the smart phone (AliveECG) transforms the electrical signal to ultrasound. The smart phone's microphone will receive the ultrasound signal and transmit the electrocardiogram as a wireless PDF format.
+
+
LEADS aVR, aVL, aVF
Smart phone ECG recording RIGHT
The AliveCor device is a standardized, bipolar single lead recorder. Using the principles of Einthoven's triangle, the standard ECG bipolar and unipolar limb leads can be reproduced. To record the standard bipolar limb leads; the AliveCor case was adapted with adhesive ECG tabs (Kendall, MT710 Diagnostic Tab Electrodes). These standard adhesive tabs are placed on the right and left forearms and the lower left leg just above the ankle. The adhesive tabs are used for the precordial leads (V1, V2, V3, V4, V5, V6) and are placed on the chest using the standard locations [1,2]. Using three 24 inch insulated wires with mini alligator clips, the standard bipolar (leads: I, II, III) and unipolar (aVR, aVL, aVF) limb leads and the precordial leads are recorded in successive fashion (Fig. 1). For the augmented unipolar limb leads, we used a positive lead as the exploring lead and the other two limb leads are connected to the negative pole. For the six chest leads, the positive electrode is the exploring electrode while the indifferent lead is negative and is located on the distant left leg. Using the principles of the Einthoven's triangle and Goldberger modification, the standardized AliveCor ECG can be modified to obtain a 12 lead ECG recording [2,3]. A standard 12 lead recording was made using a Mac 5500 recorder. In case #1 and case #2, the authors performed the standard 12 lead recording and used the same exact sites for the smart phone recording. In case #3, case #4 and case #5, the standard Mac 5500 12 lead ECG was performed by an ECG technician. The precordial lead electrode placements were not exactly the same sites as those used for the AliveCor recordings that were done at a different setting. Results of the 5 individual recordings The clinical characteristics of the 5 individuals in this feasibility study are described in Table 1. Case #1 is a 67 year old male with a normal echocardiogram, ejection fraction 0.55, normal chamber sizes, and normal valve function. A standard 12 lead electrocardiogram (ECG) and then a sequential AliveCor 12 lead ECG are performed using the same lead tabs on the limbs and chest.
LEFT
+-
+-
-
+
-+
LEADS V1-V6 RIGHT
LEFT V1 - V6
++ ++++
-
+
Fig. 1. AliveCor ECG recording. Panel A: Bipolar limb lead recordings (I, II, III) using Einthoven's triangle format. Panel B: Unipolar limb lead recordings (aVR, aVL, aVF) using the method of Goldberger with the focus exploring electrode as positive and the other two limbs as the negative. Panel C: Precordial leads V1–V6: with the positive exploring electrode on the chest and the negative electrode on the distant lower left leg.
Both the Mac 5500 standard 12 lead ECG and the AliveCor generated 12 lead ECG have the same interpretation. Both show sinus rhythm, normal intervals; normal axis + 60 degrees; left atrial abnormality and left ventricular hypertrophy by voltage. The AliveCor tracing is a sequential obtained ECG and demonstrates the presence of a sinus arrhythmia. In Fig. 2, there are comparisons of the two tracings. The email PDF transfer of the AliveCor version is reconstructed for printing the 12 lead. Despite the
G.A. Baquero et al. / Journal of Electrocardiology xx (2014) xxx–xxx Table 1 Clinical characteristics of the study individuals. Case
Age/sex
Clinical diagnosis
Baseline ECG
#1 #2 #3 #4 #5
67/M 55/M 23/M 70/M 31/M
APC’S, VPC’S Cardiomyopathy none Esophageal CA Endocarditis
Sinus rhythm rate 60; LVH, LAA Sinus rhythm rate 96; LBBB Sinus rhythm rate 70; normal Sinus tachycardia rate 116; RBBB Atrial fibrillation; Dextrocardia Junctional rhythm rate 71
APC’S = atrial premature complexes. VPC’S = ventricular premature complexes. LAA = left atrial abnormality. LVH = left ventricular hypertrophy. LBBB = left bundle branch block. RBBB = right bundle branch block.
reconstruction, the QRS amplitudes, axis, intervals, initial and terminal QRS forces and “P” wave and “T” wave vectors are remarkably similar. Case #2 is a 55 year old male with a tachycardia-induced cardiomyopathy secondary to atrial fibrillation with a rapid ventricular rate. He is now post direct current cardioversion and has symptomatic improvement in sinus rhythm. He has an underlying left bundle branch block. A standard 12 lead ECG and then the sequential AliveCor 12 lead ECG are performed using the same lead tabs on the limbs and chest. Both ECG tracings show sinus rhythm, normal PR and QT intervals but a wide left bundle branch block QRS complex and left atrial abnormality. The standard ECG and the
3
sequentially obtained AliveCor generated 12 lead ECG are compared in Fig. 3. The QRS amplitudes, axis, intervals, initial and terminal QRS forces and “P” wave and “T” wave vectors mimic the standard ECG recording. In the smart phone recording, precordial leads V1, V2 and V3 show higher ST elevation than in the Mac 5500 recording. Case #3 is a 23 year old male and is a competitive cyclist with a normal echocardiogram, ejection fraction 65%. A standard 12 lead ECG and the AliveCor generated ECG are performed on two different days. The chest leads position V1–V4 are the same but V5 and V6 are different. Both the standard ECG and the AliveCor ECG have the same interpretation. They show sinus rhythm, normal intervals, and a vertical axis. Both are normal ECGs. In Fig. 4, both tracings are compared. The limb leads QRS complexes amplitudes, axis, intervals and initial and terminal forces mimic each other on the standard and AliveCor tracings. The vector of the “P” wave is the same between the two. The chest leads V1, V2, V3 and V4 are quite similar but V5 and V6 are different given different positions. In the smart phone recording, precordial leads V1, V2, V3 show J point elevation and ST elevation that is not seen in the Mac 5500 recording. Case #4 is a 70 year old symptomatic male with shortness of breath. There is a prior history of coronary artery disease, right bundle branch block and now esophageal carcinoma. The standard 12 lead ECG is a pre-operation recording and
Fig. 2. Comparison of the standard and device generated 12 lead recordings for case #1: Top trace Mac 5500 12 lead ECG and below AliveCor ECG with normal standardization at 25 mm/second (S): Interpretation standard: sinus rhythm, rate 60 beats/minute (bpm), normal intervals: PR 0.17 s, QRS duration 0.08 s, QT 0.39 s, QRS axis +60 degrees. Left atrial abnormality, voltage criteria for left ventricular hypertrophy. AliveCor generated ECG from sequentially transmitted PDF files and then assembled as shown: Interpretation: sinus rhythm and sinus arrhythmia, rate varies 56–71 bpm, normal conduction intervals: PR 0.17 s, QRS duration 0.08 s, QT 0.4 s, QRS axis +55 degrees. Left atrial abnormality, voltage criteria for left ventricular hypertrophy. A comparison shows a true 12 for 12 ECG lead match in these ECGs recorded nearly simultaneously from the same electrode tabs. A comparison of the QRS, “P” wave and “T” wave amplitudes and axis is the same between the two ECGs.
4
G.A. Baquero et al. / Journal of Electrocardiology xx (2014) xxx–xxx
Fig. 3. Comparison of the standard and device generated 12 lead recordings for case #2: Top trace standard ECG and below device generated ECG: Interpretation standard: sinus rhythm rate 94 bpm; PR 0.16 s, QRS duration 0.17 s, QT 0.38 s, QRS axis −15 degrees; left atrial abnormality; left bundle branch block with repolarization changes. Device generated ECG from sequentially transmitted PDF files and reassembled as shown: Interpretation: sinus rhythm, rate 100 bpm; PR 0.16 s; QRS duration 0.16 s; QT 0.4 s; QRS axis −15 degrees. Left atrial abnormality; left bundle branch block with repolarization changes. The comparison shows a true 11 for 12 ECG lead match in these ECGs recorded sequentially using the same limb and precordial position electrode tabs. Only lead aVF is slightly different.
Fig. 4. Comparison of the standard and device generated 12 lead recordings for case #3: (same format as above): Interpretation standard: sinus rhythm and sinus arrhythmia rate varies from 62 bpm to 88 bpm, PR 0.12 s, QRS duration 0.11 s, QT 0.4 s, QRS axis +100 degrees. Normal ECG. Device generated ECG interpretation: sinus rhythm rate 48 bpm, PR 0.12 s, QRS duration 0.10 s, QT 0.4 s, QRS axis +100 degrees. Normal ECG. In this comparison, the limb leads from the two formats mimic each other. The chest leads, V1, V2, V3 and V4 are the same but leads V5 and V6 are different given the lead placements are different.
G.A. Baquero et al. / Journal of Electrocardiology xx (2014) xxx–xxx
5
Fig. 5. Comparison of the standard and device generated 12 lead recordings for case #4: (same format as above): Interpretation standard: sinus tachycardia, rate 116 bpm, PR 0.15 s, QRS duration 0.16 s, QT 0.38 s, QRS axis − 45 degrees. Left axis deviation; right bundle branch block with repolarization abnormality; inferior myocardial infarction, age undetermined. Device generated ECG interpretation: sinus rhythm rate 75 bpm, PR 0.18 s, QRS duration 0.13 s, QT 0.4 s, QRS axis − 45 degrees. Left axis deviation, right bundle branch block with repolarization abnormality, inferior myocardial infarction. Despite the different chest lead positions and the different days for each, the two are quite similar in most leads.
the AliveCor generated 12 lead is a postoperative recording with different chest positions. The standard 12 lead ECG shows sinus tachycardia, rate 116 bpm; wide QRS duration, normal QT, left axis deviation, − 45 degrees, right bundle branch block with a repolarization abnormality and an inferior myocardial infarction, age undetermined (Fig. 5). The AliveCor generated 12 lead ECG also shows sinus rhythm, left axis deviation, right bundle branch block and the age undetermined inferior infarct. Although the tracings are from different days, there is a remarkable similarity between the two recordings QRS durations, amplitudes and axis. The initial and terminal QRS forces and “P” wave vectors are also similar. There are ST depressions that seem exaggerated in V1 and V2 in the smart phone recording. Case #5 is a 31 year old male symptomatic with fever, malaise, shortness of breath, and a new diagnosis of endocarditis. He has a long-standing permanent pacemaker with a trans-venous right atrial pacing electrode and an epicardial ventricular pacing lead. The base line cardiac diagnoses are persistent atrial fibrillation, complete atrialventricular block and congenitally corrected transposition of the great arteries with dextrocardia. The standard 12 lead ECG and the AliveCor generated 12 lead are from different days. The standard 12 lead ECG shows underlying atrial fibrillation, an accelerated junctional rhythm at 72 bpm and dextrocardia with right axis deviation and poor R wave progression. In Fig. 6, the AliveCor generated 12 lead ECG shows atrial fibrillation, an accelerated junctional rhythm
and a comparable QRS morphology that also supports the diagnosis of dextrocardia. In the smart phone recording, there are repolarization differences in leads I, II, V1, V2 and V4–V6.
Discussion The standard 12-lead electrocardiogram remains the backbone for clinical decision-making in the acute care arena particularly for diagnosis and treatment of the acute coronary syndromes. It remains the primary tool for the recognition of and the diagnosis of cardiac arrhythmia. In both cases, the earlier the 12 lead ECG can be obtained, the earlier the diagnosis, initiation of therapy and the reduction in morbidity and mortality. The new technology is moving medicine in that direction. The AliveCor single channel validated and standardized ECG recorder for smart phone use is a step in the direction of better technology. We hypothesize from our 5 case series that the device can be easily modified to become a small, versatile and practical 12 lead ECG recorder. The smart phone is a ubiquitous part of life today and for it to have the potential to easily record a simple important medical test such as the 12 lead ECG may be potentially lifesaving. In this report, we show the feasibility of converting a smart phone based bipolar heart rhythm recorder to a 12 lead recorder with standard ECG tabs and wires with alligator
6
G.A. Baquero et al. / Journal of Electrocardiology xx (2014) xxx–xxx
Fig. 6. Comparison of the standard and device generated 12 lead recordings for case #5: (same format as above): Interpretation standard: underlying atrial fibrillation, a regular wide QRS with a QRS duration 0.12 s (non-specific intra-ventricular conduction delay), QT 0.38 s, QRS axis +150 degrees. An accelerated junctional rhythm at 72 bpm. There is right axis deviation, poor R wave progression and reduced QRS voltages in the left lateral precordial leads that are consistent with dextrocardia. Device generated ECG interpretation: Underlying atrial fibrillation, regular wide QRS rhythm, QRS duration 0.12 s, QT 0.44 s, QRS axis +150 degrees. Accelerated junctional rhythm, at 64 bpm, right axis deviation, poor R wave progression without transition consistent with dextrocardia. The ECGs are quite comparable despite different precordial lead placement and on different days.
clips from a hardware store. We simply followed the tenets of Einthoven's triangle to sequentially perform a standardized 12 lead ECG in 5 individuals. The bipolar limb lead tracings (I, II, III) are easily recorded with the AliveCor hardware. Using the semi-direct lead hypothesis of electrocardiography, we could use an “exploring electrode” on the chest near the heart's surface and a distant “indifferent electrode” on the lower left leg to record the unipolar chest leads (V1, V2, V3, V4, V5, V6) [1]. In the absence of the “Wilson central terminal,” we recorded the unipolar limb leads (aVR, aVL, aVF) with an exploring electrode on the focused limb and the other two limbs as the composite indifferent electrode. This is the Goldberger unipolar limb lead arrangement [3]. We show five examples of 12 lead ECGs recorded with the use of a smart phone and the AliveCor technology. In its present form, the AliveCor technology is different from the Mac 5500 technology. The most obvious is the ease at which the ECG is obtained on a 3-channel recorder and printed in a standard 12 lead format. The AliveCor recorder is a single channel, bipolar platform that we adapted to construct a 12 lead tracing. The AliveCor does not have a “Wilson central terminal” which is a standard feature of the Mac 5500 recorder. This feature allows for effectively separating a bipolar signal into two components and allows for the unipolar recordings. The central terminal stabilizes the
electrical forces at the indifferent electrode by creating a “zero potential”. This allows the unipolar exploring electrode to accurately depict the focal electrical activity without the interference of even a distant indifferent lead [1,2]. The AliveCor presently does not have this feature. This probably explains the differences in the magnitude of some of the repolarization differences in the smart phone recordings. These differences are magnified in the unipolar precordial leads. This is seen in case #2 (simultaneous recordings) where there is ST elevation in the precordial leads that is absent in the Mac 5500 recording. Most differences in repolarization (i.e. ST-T abnormalities) reflect the fact that in cases #3, #4 and #5, the ECG recordings were done on different days and the precordial leads are placed differently. When it comes to depolarization (QRS morphology), the accuracy of the smart phone recording appears to mimic the standard tracings in all aspects. This is particularly so in the two individuals (cases #1 and #2) where the recordings were sequential using the same leads. In the other three individuals the limb leads mimic the standard recordings but the chest leads have some differences possibly due to differing chest lead positions at the time of the two recordings on different days. This series includes three individuals with abnormal QRS complex morphology: LBBB, RBBB and dextrocardia. All three are accurately depicted by the smart phone version of the ECG. Depolarization is a shorter and less complex
G.A. Baquero et al. / Journal of Electrocardiology xx (2014) xxx–xxx
electrical event than repolarization [1]. Despite the lack of a “Wilson central terminal”, depolarization seems to be accurately replicated by the smart phone. Repolarization is prolonged and complex and it is easily influenced by any change in positional, hemodynamic, metabolic and physiological factors [1]. The populations that may find this “mobile 12 lead ECG” useful might include those patients with a prior myocardial infarction and those with documented arrhythmias. Patients with known coronary disease may develop a clinical event remote from a hospital but could record and transmit a 12 lead ECG. Those with known arrhythmias who develop symptoms can record and transmit not just the single lead event monitor strip but also a 12 lead of the tachycardia for diagnostic review. The populations that may benefit the most are in the third world where resources are limited but a smart phone is available. A smart phone 12 lead ECG in remote 3rd world locations could revolutionize the recognition of acute cardiac issues in those without access to healthcare in a more traditional setting. Despite the obvious limitations of different precordial lead positions in some individuals, the study demonstrates the feasibility of creating a 12 lead ECG with your smart phone. A minor limitation is the time it takes to construct the 12 lead recording on the smart phone. It takes us about 10– 12 minutes to obtain the 12 single channel strips and place them into 12 PDF files. This time restraint may be reduced by 50% once the learning curve has been eliminated. The tracings are easily visualized on the smart phone's screen separately. They can be easily transmitted as separate PDF files for review on another device and then printed in separate strips. The quality of the recordings would seem to
7
suggest that this technology could become an important useful tool for clinical use. The series of 5 cases is small but each supports the concept that the smart phone technology with its virtual 12 lead ECG can reliably recreate a reliable QRS complex. This does not validate the concept but merely presents it as plausible. Design of a new smart phone case (hard ware upgrade) seems a small challenge. The APPs and software upgrades to make the leads acquisitions simultaneous and easily transferable seem very plausible. The addition of a “Wilson central terminal” is also plausible and may help with stabilization of repolarization on the unipolar recordings. This new hand held smart phone 12 lead ECG recorder needs further development and validation. References [1] Lipman BS, Massie E, Kleiger RE. Clinical scalar electrocardiography. Chicago: Year Book Medical Publishers, Inc.; 197322–31. [2] Marriott HJL. Practical electrocardiography. Baltimore/London: Williams and Wilkins; 19801–7. [3] Goldberger E. Simple electrocardiographic electrode of zero potential and technic of obtaining augmented unipolar extremity leads. Am Heart J 1942;23:483–92. [4] Saxon LA, Smith A, Doshi S, Albert D. I phone rhythm strip. The implications of wireless and ubiquitous heart rate monitoring. J Am Coll Cardiol 2012;59(13s1):E726-E726. [5] Saxon LA. Ubiquitous wireless ECG recording: powerful tool physicians should embrace. J Cardiovasc Electrophysiol 2013;24 (4):480–3. [6] Mittal S, Movsowitz C, Steinberg JS. Ambulatory external electrocardiographic monitoring: focus on atrial fibrillation. J Am Coll Cardiol 2011;58(17):1741–9. [7] Kamel H, Smith WS. Detection of atrial fibrillation and secondary stroke prevention using telemetry and ambulatory cardiac monitoring. Curr Atheroscler Rep 2011;13(4):338–43.
