Access to this full-text is provided by Springer Nature.
Content available from Scientific Reports
This content is subject to copyright. Terms and conditions apply.
1
SCIENTIFIC REPORTS | (2019) 9:14497 | https://doi.org/10.1038/s41598-019-50487-4
www.nature.com/scientificreports
Utility of a Smartphone Based
System (cvrPhone) to Predict Short-
term Arrhythmia Susceptibility
Kwanghyun Sohn1, Steven P. Dalvin1, Faisal M. Merchant2, Kanchan Kulkarni
1,
Furrukh Sana1, Shady Abohashem
1, Jagmeet P. Singh4, E. Kevin Heist4, Chris Owen5,
Eric M. Isselbacher6 & Antonis A. Armoundas1,3
Repolarization alternans (RA) has been implicated in the pathogenesis of ventricular arrhythmias
and sudden cardiac death. We developed a 12-lead, blue-tooth/Smart-Phone (Android) based
electrocardiogram (ECG) acquisition and monitoring system (cvrPhone), and an application to estimate
RA, in real-time. In in-vivo swine studies (N = 17), 12-lead ECG signals were recorded at baseline and
following coronary artery occlusion. RA was estimated using the Fast Fourier Transform (FFT) method
using a custom developed algorithm in JAVA. Underlying ischemia was detected using a custom
developed ischemic index. RA from each lead showed a signicant (p < 0.05) increase within 1 min of
occlusion compared to baseline (n = 29). Following myocardial infarction, spontaneous ventricular
tachycardia episodes (n = 4) were preceded by signicant (p < 0.05) increase of RA prior to the onset
of the tachy-arrhythmias. Similarly, the ischemic index exhibited a signicant increase following
myocardial infarction (p < 0.05) and preceding a tachy-arrhythmic event. In conclusion, RA can be
eectively estimated using surface lead electrocardiograms by analyzing beat-to-beat variability in ECG
morphology using a smartphone based platform. cvrPhone can be used to detect myocardial ischemia
and arrhythmia susceptibility using a user-friendly, clinically acceptable, mobile platform.
Electrocardiographic (ECG) alternans, a phenomenon of beat-to-beat oscillation in electrocardiographic wave-
forms during the repolarization phase of the cardiac cycle also known as repolarization alternans (RA), has been
demonstrated to be an important marker of cardiac electrical instability and ventricular tachy-arrhythmic events
(VTE)1,2. Specically, the presence of microvolt level RA during low level exercise has been identied as a marker
of ventricular arrhythmia susceptibility and can be used to guide implantable cardioverter debrillator (ICD)
therapy in patients with structural heart disease.
However, beyond a risk stratication marker for patients that are candidates to receive ICD therapy, recent
clinical studies have also indicated that elevated levels of RA may have important predictive signicance of
short-term arrhythmia susceptibility. Analysis of body-surface ECG signals from ambulatory patients (Holter
monitors) with coronary artery disease has demonstrated a sharp surge in the magnitude of RA within minutes
prior to spontaneous VTEs3. Analysis of intra-cardiac electrograms (EGMs) from ICDs has demonstrated a sharp
elevation in RA magnitude immediately prior to spontaneous ventricular arrhythmias4,5. However, a similar surge
in RA has not been observed prior to induced VTEs or preceding inappropriate ICD discharges5,6. Overall, there
is signicant evidence to support the notion that a heightened state of RA, measured from intra-cardiac elec-
trodes or body-surface leads, is closely associated with an increased risk to a VTE.
On the other hand, as the average age of the US population increases and chronic conditions are becoming
more prevalent, there is a need to improve the eectiveness of disease prevention, to enhance access to healthcare,
and to sustain healthy independent living. e increased availability of new technologies and an ever-improving
health information technology infrastructure, with >90% of American adults owning a cell phone and 55%
having a Smart-Phone7, indicates that mobile-health technologies will soon function not only as monitoring
1Cardiovascular Research Center, Massachusetts General Hospital, Boston, MA, USA. 2Cardiology Division, Emory,
University School of Medicine, Atlanta, GA, USA. 3Institute for Medical Engineering and Science, Massachusetts
Institute of Technology Cambridge, MA, USA. 4Cardiology Division, Cardiac Arrhythmia Service, Massachusetts
General Hospital, Boston, MA, USA. 5Neurosurgery Division, Massachusetts General Hospital, Boston, MA, USA.
6Healthcare Transformation Lab, Massachusetts General Hospital, Boston, MA, USA. Correspondence and requests
for materials should be addressed to A.A.A. (email: aarmoundas@partners.org)
Received: 7 April 2019
Accepted: 10 September 2019
Published: xx xx xxxx
OPEN
Content courtesy of Springer Nature, terms of use apply. Rights reserved
2
SCIENTIFIC REPORTS | (2019) 9:14497 | https://doi.org/10.1038/s41598-019-50487-4
www.nature.com/scientificreports
www.nature.com/scientificreports/
devices of the cardiac and respiratory systems8, but as essential components in managing patients. erefore, new,
low-cost, easy-to-deploy technologies are needed to meet the clinical need for long-term (>1–2 days) respiratory
and cardiac monitoring of the ambulatory patient. e central goal of this study is to investigate the hypoth-
esis that one may develop methods for estimating RA, by recording cardiac electrical activity from the body
surface, measuring the beat-to-beat variability in the morphology of ECG waveforms, and using the measured
beat-to-beat variability to estimate the RA using the on-board computing power of a Smart-Phone, in order to
alert the patient and the treating physician of an impending arrhythmia.
Methods
Animal studies. 17 male Yorkshire swine (40–45 kg) were anesthetized and instrumented in the Animal
Electrophysiology Laboratory of the Massachusetts General Hospital, following previously described methods9.
Anesthesia was maintained with Isourane (1.5–5%), and each animal was intubated and was mechanically ven-
tilated. Ιnvasive blood pressure was monitored through an arterial line.
Briey, percutaneous vascular access was obtained in the jugular veins and femoral arteries and veins, as pre-
viously described, using standard Seldinger techniques10. Decapolar catheters were placed in the coronary sinus
(CS), right ventricle (RV), right atrium (RA), and le ventricle (LV). An inferior vena cava catheter was inserted
as a reference electrode for unipolar signals.
Percutaneous techniques were used to induce coronary artery ischemia, in a closed-chest model9,11–14. Briey,
either the mid le circumex or the mid le anterior descending coronary arteries were occluded with a balloon
using standard angioplasty techniques. Ischemia was validated and conrmed by hand injections of contrast
into the coronary in which case no-ow, or manifestation of ECG changes were indications of full occlusion.
Intravenous unfractionated heparin was administered (4000 units prior to engaging the coronary artery, followed
by 1000 units/hour during balloon ination).
The hardware architecture. e hardware architecture of the system has been previously described8.
Briey, the ECG device is composed of an analog-to-digital (A/D) converter, a microcontroller board, and
a Bluetooth module (Fig.1A). Following amplication and digitization of the analog ECG signal by the AD
converter, they are transmitted by the microcontroller to the smartphone at the user’s request (Fig.1B). We
have validated that signals can be uninterruptedly communicated through the Bluetooth, up to 10 m away
from the smartphone, at a baud rate of 115200. e microcontroller was programmed using the open-source,
Arduino 1.5.4.
e settings of the AD converter were: sampling rate at 500 samples/s, gain at 12 and reference voltage at
24 V. Reference voltage for the precordial leads was the Wilson Central Terminal dened as RA + LA + LL)/3).
Although, the AD converter has 24 bit resolution, that was reduced to 16 bit in order to reduce the transmission
load via Bluetooth. e range of the ECG signal is ±12.5 mV, and its resolution is ~0.38 μV.
Android smartphone application. e application is consisted of three threads: the user-interface, the
Bluetooth, and the real-time-calculation. e user is provided with diverse options through the user-interface
thread, such as to display the ECG signals and the estimation results. e Bluetooth thread receives the ECG
signals from the microcontroller. e real-time-calculation thread estimates RA indices for each lead, inde-
pendently, and in real-time.
Body surface ecg data analysis. RA is estimated using a previously described algorithm2,4,9,15. Briey, we
rst obtain preliminary R-wave detection by applying a soware-based QRS detection algorithm to a selected
ECG lead. ese, preliminary R-wave detections are rened and abnormal beats (i.e. premature ventricular com-
plexes -PVCs- and aberrantly conducted beats) are identied by employing a template-matching QRS alignment
algorithm and substituted with a median odd or even template beat (estimated from the odd or even ‘normal’
beats respectively in the 128 beat sequence), depending on whether the abnormal beat is an odd or an even beat2,9
en, repolarization interval boundaries for RA analysis are independently determined for each of the body
surface leads, due to variability in the morphology and timing of the T-wave between leads. Briey, the power
method identies the onset/oset points at time points corresponding to 5% and 95% of the cumulative sum of
the signal power16, is used for ECG signal waveform annotation.
e, RA is estimated using the spectral method for each 128-beat data sequence (using a 512-point power
spectrum to improve the frequency-domain resolution), as previously described2,9,15,17. For each lead, spectral
analysis is independently performed in order to account for the spatial variability of RA, and RA indices are
estimated as follows:
μ= −μ
Valternansvoltage () alternanspeaknoise
=
−μ
σ
K
alternanspeak
scorenoise
noise
where, the alternans peak is the peak in the aggregate power spectrum corresponding to 0.5 cycles/beat and the
mean (µnoise) and the standard deviation (σnoise) of the alternans noise are estimated in a predened spectral win-
dow (0.43–0.46 cycles/beat) of the power spectrum. e alternans voltage measures directly the presence of RA
while the Kscore is a measure of the statistical signicance of the alternans voltage. For each lead, RA is estimated
on a beat-by-beat basis using a rolling 128-beat window that is shied one beat at a time.
Content courtesy of Springer Nature, terms of use apply. Rights reserved
3
SCIENTIFIC REPORTS | (2019) 9:14497 | https://doi.org/10.1038/s41598-019-50487-4
www.nature.com/scientificreports
www.nature.com/scientificreports/
Ischemic index estimation. ST-segment elevation or depression has been well established as a signicant
marker of MI18. We have previously introduced the ischemic index19, which is dened as the absolute value of the
ratio of ST-height to the QR-amplitude. e ST-height is dened as the mean amplitude of the whole ST-segment
above or below the isoelectric baseline, when the polarity at both ends of the ST-segment is the same; if the polar-
ity is dierent, then the longer segment is selected as the ST-height.
Assessment of arrhythmia susceptibility. Arrhythmia susceptibility, under varying states of RA,
was assessed using programmed ventricular stimulation (PVS)20, in which a positive outcome was dened as
sustained ventricular tachycardia (VT) or ventricular brillation (VF) lasting >30 secs or requiring external
debrillation.
Pacing pulses during PVS were delivered from LV15 and had amplitude and duration 50 mA and 2 msec,
respectively. PVS was initiated with a drive train of 8 beats (S1) at a cycle length of 400 milliseconds (ms) with
an extra-stimulus (S2) delivered at a coupling interval of approximately 300 ms. e coupling interval for S2 was
reduced in 10 ms steps until ventricular refractoriness was reached, at which point S2 was xed at 20 ms above the
point of refractoriness and an S3 was added beginning at a coupling interval 10 ms less than S2. is process was
repeated until sustained VT/VF was induced or ventricular refractoriness was reached on S6, in which case PVS
was deemed non-inducible under those conditions.
In order to quantify the outcomes of PVS across dierent RA states, we developed a single “score” rank param-
eter (Srank) which assigned the highest score (highest arrhythmia susceptibility) to the intervention that required
(i) the smallest number of extra-stimuli during PVS to induce an arrhythmia, or (ii) if the number of extra-stimuli
was the same, to the intervention with the smallest coupling interval between S1 and Slast, both of which suggest
less aggressive stimulation was necessary to induce sustained VT/VF reecting a more vulnerable arrhythmic
substrate. We recognize that there is no single best validated clinical method to assess arrhythmia susceptibility
Figure 1. e smartphone-based repolarization alternans monitoring system. e Bluetooth-enabled ECG
acquisition device is composed of three parts: An analog-to-digital (AD) converter, a microcontroller board,
and a Bluetooth module. e AD converter amplies and digitizes the signals from the ten electrodes on the
torso, and the microcontroller transmits the signals to the smartphone through the Bluetooth module. en, the
smartphone calculates repolarization alternans indices for each lead in real-time.
Content courtesy of Springer Nature, terms of use apply. Rights reserved
4
SCIENTIFIC REPORTS | (2019) 9:14497 | https://doi.org/10.1038/s41598-019-50487-4
www.nature.com/scientificreports
www.nature.com/scientificreports/
in a fully quantiable manner. e Srank score was developed not as surrogate of VT/VF (with a binary outcome),
but rather as a method to obtain a quantitative relationship between the level of RA and the likelihood of inducing
VT/VF.
Figure 2. Coronary artery occlusion induced temporal changes of the estimated repolarization alternans (ST-
segment and T-wave) indices (n = 29 records; N = 17 animals): (A) alternans noise (µnoise), (B) alternans voltage,
and (C) Kscore. Time zero indicates the balloon ination moment. Each bar graph represents 10, 25, 50, 75 and
90 percentiles of the corresponding alternans index values beat-by-beat estimated for all animals for 1 minute
time span. Asterisk indicates statistically signicant increase aer occlusion, compared to before occlusion
(p < 0.0001 for the alternans noise, p < 0.0001 for the alternans voltage and p < 0.05 for the Kscore).
Content courtesy of Springer Nature, terms of use apply. Rights reserved
5
SCIENTIFIC REPORTS | (2019) 9:14497 | https://doi.org/10.1038/s41598-019-50487-4
www.nature.com/scientificreports
www.nature.com/scientificreports/
If sustained VT/VF was induced, biphasic external debrillation was performed using 150 joules with paddles
placed on the chest of the animal and a rest period of ~10 min was allowed aer each positive PVS.
Statistical methods. Aggregate variables are expressed as mean ± standard deviation. Box-plot representa-
tion including the median, 90–10% and 75–25% percentiles was used to demonstrate statistical properties of the
estimated data sequences. For each RA parameter, a baseline distribution was obtained by collecting the values
of that parameter over all time periods before occlusion (t < = 0). Comparisons were then made for each of the
alternans noise (µnoise), alternans voltage and Kscore, for each lead, between the baseline distribution and the distri-
bution corresponding to each minute aer occlusion (t > 0), and a p value was obtained using the Kruskal Wallis
test. A threshold value of 0.05 divided by the number of time intervals aer occlusion was calculated. Statistical
signicance at any time interval was then determined based on two factors: (i) the p value resulting from the com-
parison between the baseline distribution with distribution at that particular interval is less than the threshold
value, and (ii) the median of the baseline distribution is less than the median of the distribution at that particular
interval. A statistically signicant p value is denotted by an “*. Statistical analysis was performed using MATLAB
(MathWorks Inc, Natick, MA).
Ethical approval. e animal studies were approved by the institutional review board and the subcommittee
on research animal care at Massachusetts General Hospital. All experiments were performed in accordance with
relevant guidelines and regulations.
Results
Smartphone-based repolarization alternans estimation. In Fig.2, we observe summary results
(n = 29 records, N = 17 animals) of coronary artery occlusion induced temporal changes of the estimated RA
(that involves both the ST-segment and T-wave) indices: (A) alternans noise (µnoise), (B) alternans voltage, and
(C) Kscore. Time zero indicates the timing of the balloon ination. Across all 12 ECG leads a signicant change
(p < 0.05) of the alternans noise (in a few leads), voltage and Kscore aer occlusion, compared to before occlusion,
is observed.
Repolarization alternans before a tachy-arrhythmic event. In Fig.3, we present a sample ECG signal
(lead V3) during coronary artery occlusion, while the heart-rhythm transitions from sinus to VT. In Fig.4A–C,
we observe summary results of the alternans indices following myocardial infarction, reecting temporal changes
that led led to spontaneous VT/VF (n = 4 records; N = 4 animals): (A) alternans noise (µnoise), (B) alternans volt-
age, and (C) Kscore. Time zero indicates the timing of the balloon ination. We observe that the alternans noise
level was statistically dierent (p < 0.05) before compared to aer occlusion, and also ischemia led to a statistically
signicant increase of the alternans voltage (p < 0.05) and Kscore (p < 0.05) aer occlusion, compared to before
occlusion.
We compared distributions of alternans noise (µnoise), alternans voltage, and Kscore, between records that exhib-
ited VT/VF (n = 4) and those that did not (n = 25), following myocardial infarction (at times: 0, 1, 2, 3 and 4 min),
and we report the obtained range of p-values, resulting from this comparison, in Table1.
To examine the sensitivity of the 12 lead system in detecting RA we calculated the conditional probability
that any one lead in a combination of N leads is positive, given that at least one lead out of all 12 leads is positive:
P(any one in N leads is positive | one of 12 leads is positive). We dene as positive RA an estimate that satises the
following criteria: (i) alternans voltage is higher than 0.55 μV, and (ii) Kscore is higher than 39. If at any instance, we
nd that any one of the 12 leads is positive, we evaluate if positive alternans can be detected with a combination of
N leads, with N ranging from one to twelve. All combinations of N leads out of 12 have been considered for this
purpose. en, the probability for a specic combination of leads is calculated by the ratio between the number
of times a positive detection was made to the total number of positive detections by the 12 leads. Once the prob-
abilities are computed over all combinations of size N across all 29 recordings, the average probability over the
Figure 3. ECG signal (lead V3) displaying spontaneous transition to ventricular tachycardia aer coronary
artery occlusion.
Content courtesy of Springer Nature, terms of use apply. Rights reserved
6
SCIENTIFIC REPORTS | (2019) 9:14497 | https://doi.org/10.1038/s41598-019-50487-4
www.nature.com/scientificreports
www.nature.com/scientificreports/
29 recordings for each combination was calculated, and the combination yielding the maximum probability for
a specic number of leads was reported (Fig.5). We observe that four leads provide higher than 80% probability
that RA is detected and that number raised to more than 90% with six leads.
Figure 4. Temporal changes of the repolarization alternans (ST-segment and T-wave) indices during
myocardial infarction that led to spontaneous ventricular tachycardia/brillation (n = 4 records; 4 animals):
(A) alternans noise (µnoise), (B) alternans voltage, and (C) Kscore. Time zero indicates the balloon ination
moment. Each bar graph represents 10, 25, 50, 75 and 90 percentiles of the corresponding alternans index
estimated on a beat-by-beat basis for all animals, in 1 min time intervals. e asterisk indicates a statistically
signicant increase aer occlusion compared to before occlusion (p < 0.05 for the alternans noise, p < 0.05 for
the alternans voltage and p < 0.05 for the Kscore).
Content courtesy of Springer Nature, terms of use apply. Rights reserved
7
SCIENTIFIC REPORTS | (2019) 9:14497 | https://doi.org/10.1038/s41598-019-50487-4
www.nature.com/scientificreports
www.nature.com/scientificreports/
Repolarization alternans burden. In Fig.6, we present the alternans burden (%) before and aer coro-
nary artery occlusion during MI (n = 29 records; N = 17 animals). Again, we dene as positive an RA an estimate
that satises the criteria above for (i) alternans voltage is higher than 0.55 μV, and (ii) Kscore is higher than 39.
e incidence of RA is evaluated on a beat-by-beat basis, and the RA burden is evaluated as a percent of
sequences that exhibit signicant RA; the RA burden is estimated separately aer the occlusion, for each record.
We observe that during MI the RA burden is signicantly higher (p < 0.05, using the paired T-test), compared
to baseline.
Relationship of ischemic index and repolarization alternans. Next, we sought to explore the rela-
tionship of RA vs the ischemic index during MI (Fig.7A) and preceding VT/VF (Fig.7B). In each gure, the
alternans voltage (μV) versus ischemic index is presented in the upper panel, and the Kscore versus ischemic index,
is presented in the lower panel. e color bars on the right side indicate the time aer coronary artery occlusion
from 0 min to 5 min. e dashed line in each plot represents a data tting line with a single-term exponential
model.
In Fig.7C, we observe that for both the alternans voltage (p < 0.05) and Kscore (p < 0.05, using the paired t-test)
the constant of the exponential model is signicantly smaller before VT/VF, indicating that RA manifests a pro-
found arrhythmogenic substrate.
RA and arrhythmia susceptibility. To assess the arrhythmogenic potential of RA we employed PVS that
was performed at baseline and aer coronary artery occlusion (N = 9).
We observed that the Srank at baseline and after coronary artery occlusion was not statistically different
(Fig.7D), yet it trended towards a higher value aer occlusion associating RA with a higher arrhythmogenic risk.
Lead Alternans Noise Alternans
Voltage Kscore
I 0 < P < 0.001 0 < P < 0.001 0 < P < 0.001
II 0 < P < 0.001 0.001 < P < 0.733 0.001 < P < 0.257
III 0 < P < 0.186 0.001 < P < 0.843 0.001 < P < 0.068
AVR 0 < P < 0.001 0.001 < P < 0.492 0.001 < P < 0.088
AVL 0 < P < 0.015 0.001 < P < 0.362 0.001 < P < 0.776
AVF 0.001 < P < 0.090 0.001 < P < 0.429 0.001 < P < 0.007
V1 0.001 < P < 0.003 0.001 < P < 0.777 0.001 < P < 0.944
V2 0.001 < P < 0.423 0.001 < P < 0.318 0.001 < P < 0.783
V3 0.001 < P < 0.124 0.001 < P < 0.186 0.001 < P < 0.418
V4 0.001 < P < 0.098 0 < P < 0.001 0 < P < 0.001
V5 0.001 < P < 0.007 0.001 < P < 0.754 0.001 < P < 0.713
V6 0 < P < 0.001 0 < P < 0.001 0.001 < P < 0.243
Table 1. Range of p-values resulting from comparing distributions of alternans noise, alternans voltage, and
Kscore, between records that exhibited VT/VF (n = 4) and those that did not (n = 25), following myocardial
infarction (at times: 0, 1, 2, 3 and 4 min, in Figs2 and 3).
Figure 5. Sensitivity of the 12 lead ECG in detecting RA, that is P(any one in N leads is +| one of 12 leads is +).
In the plot one observes the highest performing lead combinations of N leads, for any number of leads ranging
from one to twelve. +: indicates positive.
Content courtesy of Springer Nature, terms of use apply. Rights reserved
8
SCIENTIFIC REPORTS | (2019) 9:14497 | https://doi.org/10.1038/s41598-019-50487-4
www.nature.com/scientificreports
www.nature.com/scientificreports/
Discussion
In this study, we have shown that RA can be eectively estimated from body surface ECG signals, through
Bluetooth, using a smartphone; second, the smartphone can provide a viable platform to process ECG signals in
real-time and, if needed, enable generation of alerts for the patient and the treating physician of an impending
arrhythmia while the patient maintains an ambulatory status; third, there is a strong connection between RA and
the ischemic index, especially before a tachy-arrhythmic event, indicating the signicance of RA in predicting a
tachy-arrhythmic event, at least in this model.
Optical mapping studies in normal hearts have shown that discordant (reecting two areas in the heart that
oscillate with opposing phase) APD alternans is linked to a state of reduced cardiac electrical stability, manifested
by the observation that when alternans is followed by VF, it only occurs aer discordant APD alternans, but never
concordant APD alternans21.
RA estimated in Holter ECG signals in ambulatory patients with coronary artery disease has shown a marked
surge in RA magnitude within minutes preceding a spontaneous VTE3. T-wave alternans (TWA) amplitude
reached a peak about 10 min prior to the onset of a VTE. Sharp surges in TWA immediately preceding sponta-
neous VTEs have also been documented in body-surface ECGs in patients hospitalized for acute heart failure22;
TWA increased from a baseline during 15–30 mins prior to the onset of the VTE and remained elevated until the
occurrence of VTE. RA estimated in intra-cardiac EGMs from ICDs has shown a sharp surge prior to spontane-
ous VTEs4,5; however, a similar RA surge has not been noticed prior to induced VTEs or prior to inappropriate
ICD shocks5. Recently a prospective study in patients with ICDs has conrmed these ndings6; specically, the
magnitude of T-wave alternans/variability (TWA/V) prior to spontaneous VTE was signicantly higher than dur-
ing any of the control segments, while logistic regression analysis has shown that each 10 μV increase in TWA/V
was associated with a 2.2 odds increase of developing a VTE. ese observations establish a close temporal rela-
tionship between surges in TWA/V and the onset of spontaneous VTEs.
On the other hand, the ischemic index, that quanties beat-to-beat changes observed in both ventricular
depolarization and repolarization during ischemia, provides a personalized, lead-independent measure that
accounts for both depolarization23,24 and repolarization25–27 changes observed during MI. In this study, as well as
in prior studies28, we have seen that despite the dynamic beat-to-beat and subject-to-subject variability of ECG
morphology, the ischemic index presents high stability as well as very low intra- and inter-subject variability
under baseline (non-ischemic) conditions28, while it exhibits great spatial sensitivity in detecting MI-induced
changes and has been linked to VTEs28.
Figure 6. Repolarization alternans (ST-segment and T-wave, RA) burden before and aer coronary artery
occlusion. RA positive, criteria were dened as: (i) alternans voltage is greater than 0.55 μV, and (ii) Kscore greater
than 3. e RA burden is evaluated on a beat-by-beat basis as a percent of sequences that exhibited signicant
RA, and percentages of RA incidence are calculated before and aer the occlusion separately, for each record.
Each bar graph represents 10, 25, 50, 75 and 90 percentiles of alternans burden of all records. An asterisk
indicates statistically signicant (p < 0.05) dierence between the two alternans percents before and aer
occlusion (n = 29 records; N = 17 animals).
Content courtesy of Springer Nature, terms of use apply. Rights reserved
9
SCIENTIFIC REPORTS | (2019) 9:14497 | https://doi.org/10.1038/s41598-019-50487-4
www.nature.com/scientificreports
www.nature.com/scientificreports/
In summary, although the magnitude of RA increases in body-surface leads is smaller than that measured
in intra-cardiac EGMs9, simultaneous measurement of RA from body-surface and intra-cardiac EGMs by our
group9 and others29 has shown a high degree of correlation suggesting that these measurements are reecting the
same electrical phenomenon. e data presented in this study as well as by others support the idea that a sharp
Figure 7. Relationship of repolarization (ST-segment and T-wave) alternans vs ischemic index (A) during
myocardial infarction and (B) preceding ventricular tachycardia/brillation. Alternans voltage (μV) versus
ischemic index (upper panel), and Kscore versus ischemic index (lower panel). e color bars on the right side
show time aer coronary artery occlusion from 0 min to 5 min. e dashed line at each plot represents a data
tting line with a single-term exponential model. (D) PVS that was employed at baseline and aer coronary
artery occlusion, MI (N = 9), to assess the arrhythmogenic potential of RA. Although the Srank at baseline and
aer coronary artery occlusion was not statistically dierent, yet it trended towards a higher value aer occlusion.
Content courtesy of Springer Nature, terms of use apply. Rights reserved
10
SCIENTIFIC REPORTS | (2019) 9:14497 | https://doi.org/10.1038/s41598-019-50487-4
www.nature.com/scientificreports
www.nature.com/scientificreports/
increase of RA prior to the onset of spontaneous VTE can be measured from body-surface electrodes and may be
used to predict acute arrhythmia susceptibility. In such scenario, a heightened state of the ischemic index and/or
RA (compared to that subject’s baseline levels, personalized health care) could serve as a warning and indication
that the subject should adopt behavioral changes (i.e. stop exercising) or take medication (i.e. a b-blocker), or seek
medical attention.
Data Availability
e data will be available to any investigator upon request.
References
1. itzenberg, A. L., Adam, D. . & Cohen, . J. Period multupling-evidence for nonlinear behaviour of the canine heart. Nature 307,
159–161 (1984).
2. Smith, J. M., Clancy, E. A., Valeri, C. ., usin, J. N. & Cohen, . J. Electrical alternans and cardiac electrical instability. Circulation
77, 110–121 (1988).
3. Shusterman, V., Goldberg, A. & London, B. Upsurge in T-wave alternans and nonalternating repolarization instability precedes
spontaneous initiation of ventricular tachyarrhythmias in humans. Circulation 113, 2880–2887 (2006).
4. Armoundas, A. A., Albert, C. M., Cohen, . J. & Mela, T. Utility of implantable cardioverter debrillator electrograms to estimate
repolarization alternans preceding a tachyarrhythmic event. J Cardiovasc Electrophysiol 15, 594–597, https://doi.
org/10.1046/j.1540-8167.2004.03411.x (2004).
5. im, J. W. et al. Defibrillator electrogram T wave alternans as a predictor of spontaneous ventricular tachyarrhythmias in
debrillator recipients. Circ J 73, 55–62 (2009).
6. Swerdlow, C. et al. Intracardiac electrogram T-wave alternans/variability increases before spontaneous ventricular tachyarrhythmias
in implantable cardioverter-debrillator patients: a prospective, multi-center study. Circulation 123, 1052–1060 (2011).
7. Pew esearch Internet Project. Mobile technology fact sheet, http://www.pewinternet.org/fact-sheets/mobile-technology-fact-sheet
(2014).
8. Sohn, . et al. A Novel Point-of-Care Smartphone Based System for Monitoring the Cardiac and espiratory Systems. Scientic
reports 7, 44946, https://doi.org/10.1038/srep44946 (2017).
9. Weiss, E. H. et al. A novel lead conguration for optimal spatio-temporal detection of intracardiac repolarization alternans. Circ
Arrhythm Electrophysiol 4, 407–417 (2011).
10. Wilber, D., Garan, H. & usin, J. Electrophysiologic testing in survivors of cardiac arrest. Circulation 75, 146–153 (1987).
11. Merchant, F. M., Sayadi, O., Moazzami, ., Puppala, D. & Armoundas, A. A. T-wave alternans as an arrhythmic ris stratier: state
of the art. Curr Cardiol Rep 15, 398 (2013).
12. Merchant, F. M. et al. A translational approach to probe the proarrhythmic potential of cardiac alternans: a reversible overture to
arrhythmogenesis? Am J Physiol Heart Circ Physiol 306, H465–474 (2014).
13. Saul, J. P., ea, . F., Ecberg, D. L., Berger, . D. & Cohen, . J. Heart rate and muscle sympathetic nerve variability during reex
changes of autonomic activity. Am J Physiol 258, H713–721 (1990).
14. Sayadi, O. et al. A novel method for determining the phase of T-wave alternans: diagnostic and therapeutic implications. Circ
Arrhythm Electrophysiol 6, 818–826 (2013).
15. osenbaum, D. S. et al. Electrical alternans and vulnerability to ventricular arrhythmias. N Engl J Med 330, 235–241 (1994).
16. Merri, M. et al. In Proceedings. Computers in Cardiolog y 85–87 (1988).
17. Armoundas, A. A., Albert, C. M., Cohen, . J., Mela, T. & Investigators, T. Utility of implantable cardioverter defibrillator
electrograms to estimate repolarization alternans preceding a tachyarrhythmic event. Journal of cardiovascular electrophysiology 15,
594–597, https://doi.org/10.1046/j.1540-8167.2004.03411.x (2004).
18. Myerburg, . J., essler, . M. & Castellanos, A. Sudden cardiac death. Structure, function, and time-dependence of ris. Circulation
85, I2–10 (1992).
19. Zhang, L., Hou, Y. & Po, S. S. Obstructive Sleep Apnoea and Atrial Fibrillation. Arrhythm Electrophysiol Rev 4, 14–18 (2015).
20. Donahue, J. . et al. Focal modication of electrical conduction in the heart by viral gene transfer. Nat Med 6, 1395–1398 (2000).
21. Laurita, . ., Pastore, J. M. & osenbaum, D. S. In Cardiac Electrophysiology: From Cell to Bedside (eds Zipes, D. P. & Jalife, J.)
239–248 (W.B.Saunders, 1999).
22. Nearing, B. D. et al. Crescendo in depolarization and repolarization heterogeneity heralds development of ventricular tachycardia
in hospitalized patients with decompensated heart failure. Circ Arrhythm Electrophysiol 5, 84–90, https://doi.org/10.1161/
CICEP.111.965434 (2012).
23. ingborn, M. et al. Comparison of high-frequency QS components and ST-segment elevation to detect and quantify acute
myocardial ischemia. J Electrocardiol 43, 113–120 (2010).
24. omero, D., ingborn, M., Laguna, P., Pahlm, O. & Pueyo, E. Depolarization changes during acute myocardial ischemia by
evaluation of QS slopes: standard lead and vectorial approach. IEEE Trans Biomed Eng 58, 110–120 (2011).
25. Holland, . P. & Broos, H. TQ-ST segment mapping: critical review and analysis of current concepts. Am J Cardiol 40, 110–129
(1977).
26. ruco, M. W. et al. Continuously updated 12-lead ST-segment recovery analysis for myocardial infarct artery patency assessment
and its correlation with multiple simultaneous early angiographic observations. Am J Cardiol 71, 145–151 (1993).
27. Shah, A. et al. Prognostic implications of TIMI ow grade in the infarct related artery compared with continuous 12-lead ST-
segment resolution analysis. eexamining the “gold standard” for myocardial reperfusion assessment. J Am Coll Cardiol 35, 666–672
(2000).
28. Sayadi, O. et al. A novel method to capture the onset of dynamic electrocardiographic ischemic changes and its implications to
arrhythmia susceptibility. J Am Heart Assoc 3 (2014).
29. Paz, O. et al. Detection of T-wave alternans using an implantable cardioverter-debrillator. Heart Rhy thm 3, 791–797 (2006).
Acknowledgements
e work was supported by a Grand-in-Aid (#15GRNT23070001) from the American Heart Association (AHA),
the RICBAC Foundation, NIH grant 1 R01 HL135335-01, 1 R21 HL137870-01 and 1 R21EB026164-01 and a
Founders Aliate Post-doctoral Fellowship (#15POST22690003) from the AHA. is work was conducted with
support from Harvard Catalyst, e Harvard Clinical and Translational Science Center (National Center for
Research Resources and the National Center for Advancing Translational Sciences, National Institutes of Health
Award 8UL1TR000170-05 and nancial contributions from Harvard University and its aliated academic health
care centers). e content is solely the responsibility of the authors and does not necessarily represent the ocial
views of Harvard Catalyst, Harvard University and its aliated academic health care centers, or the National
Institutes of Health.
Content courtesy of Springer Nature, terms of use apply. Rights reserved
11
SCIENTIFIC REPORTS | (2019) 9:14497 | https://doi.org/10.1038/s41598-019-50487-4
www.nature.com/scientificreports
www.nature.com/scientificreports/
Author Contributions
Kwanghyun Sohn, PhD: Participated in the development of the algorithms, the animal studies, the data analysis,
writing the manuscript. Steven P. Dalvin, MD: Participated in the development of the algorithms, the data
analysis, writing the manuscript. Faisal M. Merchant, MD: Participated in the conception of the study, the
animal studies, writing the manuscript. Kanchan Kulkarni, PhD: Participated in the animal studies, writing the
manuscript. Furrukh Sana, PhD: Participated in the data analysis, writing the manuscript. Shady Abohashem,
MD: Participated in the animal studies, writing the manuscript. Jagmeet P. Singh, MD, PhD: Participated in the
conception of the study, writing the manuscript. E. Kevin Heist, MD, PhD: Participated in the conception of the
study, writing the manuscript. Chris Owen, MS: Participated in the animal studies. Eric M. Isselbacher, MD, MSc:
Participated in the conception of the study, writing the manuscript. Antonis A. Armoundas, PhD: Participated
in the conception of the study, the animal studies, the development of the algorithms, data analysis, writing the
manuscript.
Additional Information
Supplementary information accompanies this paper at https://doi.org/10.1038/s41598-019-50487-4.
Competing Interests: e authors declare no competing interests.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional aliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International
License, which permits use, sharing, adaptation, distribution and reproduction in any medium or
format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Cre-
ative Commons license, and indicate if changes were made. e images or other third party material in this
article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons license and your intended use is not per-
mitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the
copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
© e Author(s) 2019
Content courtesy of Springer Nature, terms of use apply. Rights reserved
1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com
Available via license: CC BY
Content may be subject to copyright.
Content uploaded by Shady Abohashem
Author content
All content in this area was uploaded by Shady Abohashem on Oct 11, 2019
Content may be subject to copyright.
