Abstract
Electrocardiography is an essential tool in the arsenal of medical professionals, Traditionally, patients have been required to meet health care practitioners in person to have an electrocardiogram (ECG) recorded and interpreted. This may result in paroxysmal arrhythmias being missed, as well as decreased patient convenience, and thus reduced uptake. The advent of wearable ECG devices built into consumer smartwatches has allowed unparalleled access to ECG monitoring for patients. Not only are these modern devices more portable than traditional Holter monitors, but with the addition of artificial intelligence (AI)-led rhythm interpretation, diagnostic accuracy is improved greatly when compared with conventional ECG-machine interpretation. The improved wearability may also translate into increased rates of detected arrhythmias. Despite the many positives, wearable ECG technology brings with it its own challenges. Diagnostic accuracy, managing patient expectations and limitations, and incorporating home ECG monitoring into clinical guidelines have all arisen as challenges for the modern clinician. Decentralized monitoring and patient alerts to supposed arrhythmias have the potential to increase patient anxiety and health care visitations (and therefore costs). To better obtain meaningful data from these devices, provide optimal patient care, and provide meaningful explanations to patients, providers need to understand the basic sciences underpinning these devices, how these relate to the surface ECG, and the implications in diagnostic accuracy. This review article examines the underlying physiological principles of electrocardiography, as well as examines how wearable ECGs have changed the clinical landscape today, where their limitations lie, and what clinicians can expect in the future with their increasing use.
JMIR Cardio 2025;9:e62719doi:10.2196/62719
Keywords
Introduction
The electrocardiogram (ECG) is one of the most commonly obtained test results in medical practice [,]. By measuring the electrical activity of the heart, an ECG can indicate cardiac arrhythmias and structural defects, respiratory disease, electrolyte disturbances, and even noncardiac events such as subarachnoid hemorrhage []. Traditional 12-lead ECGs are obtained by placing 10 adhesive electrodes on a patient, recording 10 seconds of electrical activity, and this snapshot is recorded for interpretation []. With the modern explosion of portable digital technology, a single lead ECG can now be performed without adhesive electrodes on a patient, using their own smart device, and these digital ECGs can be sent across vast distances for real-time clinician interpretation anywhere, at any time []. Whilst early, studies have suggested that the positive predictive value for arrhythmias such as atrial fibrillation (AF) may lie between 84% and 97% [,]. With a range of popular wearable technologies incorporating this feature, more number of patients with low cardiac risk have continuous ECG monitoring than ever before. This, plus the increasing role of deep learning and artificial intelligence (AI) in ECG interpretation, have implications for medical practitioners. More patients will be presenting with possibly abnormal ECGs recorded by their home devices, with associated anxiety and health care use already reported []. It is up to physicians have a thorough understanding of the basic sciences underpinning ECG acquisition in order to provide ECG interpretation and explain how these new devices work. This article will review the fundamentals of the ECG before examining the potential impacts of the digital age on electrocardiography for the modern doctor.
History of the ECG
This history of the ECG is really the history of electrophysiology, which can be traced back to Galvani’s [] experimentation in the 18th century on the role of electricity in the frog nervous system. More researchers followed him, and in 1902, Einthoven broke new ground by accurately recording the electrical activity of the heart using his string galvanometer [,]. The string galvanometer was not without its drawbacks; it required the patient to place their hands and 1 foot into a saltwater solution, 5 assistants to operate, and weighed over 300 kilograms [].
Thankfully, modern ECG machines have evolved, and now require only 10 small electrodes to be placed on the patient to obtain an almost complete view of the heart. Despite this, the basic principles underpinning ECG acquisition and interpretation remain unchanged since its 1902 inception, an understanding of cardiac anatomy and physiology, and physics.
The ECG: Underlying Physiological Fundamentals
Cardiomyocytes have a positive charge on their outer membrane that result from the intra- and extracellular distribution of ions. At rest, potassium (K+) ions are at a high concentration intracellularly whilst sodium (Na+), calcium (Ca2+), and chloride (Cl-) have a higher concentration outside of the cell []. The balance of ion flow (predominantly by the outward diffusion of K+ owing to membrane permeability) results in a resting membrane potential (RMP) of around −90mV []. Pacemaker cardiomyocytes have no stable RMP; instead, there is a constantly slowly increasing membrane potential mediated by the slow Na+ “funny current” (If) []. Contractile myocytes are depolarized after pacemaker cells depolarize, thereby opening If T and L-type Ca2+ channels. Fast-Na+ channels then open and allow an influx of positive Na+ ions, depolarizing the cell to about +20mV and opening slow L-type Ca2 channels. Once these channels close, active transports for sodium and calcium begin removing these ions to restore ionic equilibrium and a potassium rectifier channel will open, allowing K+ ions to leave the cell again, repolarizing the cell (
Figure 1) [,].
As each cell’s membrane becomes positively charged during depolarization, they propagate their action potentials to other nearby cells, and so on. In each wavefront of depolarization, there will be positive and negative ends, which result in a moving electrical dipole [].
A moving electrical dipole creates an electrical current. By virtue of the body’s ability to act as a volume conductor, the current field created by the flow of electricity (caused by cardiac depolarization) is conducted to the thoracic cavity, and from there, the surface of the body [,]. This current flow is thus detectable as an electrical field on the skin by surface electrodes. The 2 electrodes act as voltmeters at their respective points and measure the potential difference between them, with the “view“ between the positive and negative electrode known as a lead.For example, Lead I represents the potential difference between voltages measured at the right arm (RA; negative electrode) and left arm (LA; positive electrode) []. As an electric field moves toward the left arm (positive electrode), a positive potential difference (or voltage) is recorded, which would be reported as an upstroke in the ECG trace [].
It is important to remember that there are many thousands of myocardial fibers, each with its own electrical wavefront. Surface electrodes will not be able to distinguish the electrical field generated by each wavefront, and so, the electrical field detectable on the surface of the chest wall is determined by the vectoral sum of the electromotive field strength of all active components of the myocardium []. It is this overall vector sum (or cardiac dipole) that is represented by the ECG trace. Having multiple leads allows simultaneous recording of the same current flow in many different views. Traditionally, a 12-lead view is used in clinical electrocardiography. This includes Einthoven’s original 3-lead view, as well as 3 augmented leads (which are unipolar with a neutral central terminal) and 6 precordial leads (whose leads lie in a transverse plane) []. This requires the placement of 10 separate electrodes to create an electrical window for each lead [].
A Modern Take
Recently, breakthroughs in both the hardware and software of mobile devices have drastically changed the paradigm of ambulatory ECG monitoring, allowing ECG monitoring using wearable devices and the immediate analysis of ECGs using AI. Mobile devices are almost ubiquitous in modern society and are used daily by 2-3 billion people []. In a society where patients are eager for more involvement in their health and have a smartphone at their fingertips, it should come as little surprise that technology for home health monitoring has developed at a rapid pace. The wearable ECG device is an example of this, available using such devices as Kardia Band (AliveCor) and the Apple Watch.
The basic science principles behind these devices are the same as the traditional ECG. The device (whether it be a phone case, watch case, or other portable device) will have 2 metal plates that create the positive and negative electrodes of Lead I. When the right and left hands (or a wrist) touch both of these electrodes, a bipolar Lead I is created, as per Einthoven’s original triangle (
Figure 2) []. The signal is detected using the same principles of voltage conductance and vector analysis as the traditional ECG and interpreted using propriety AI software []. This ECG can then be stored, printed, or sent directly to physicians for interpretation and management.
Ambulatory cardiac monitoring is by no means a new development; Holter first reported the use of his eponymous cardiac monitor in 1961 [,]. However, this new hardware represents a large step forward in making it more accessible and has several advantages over the traditional Holter monitor. Whilst portable, Holter monitors are still bulky and uncomfortable to wear; they require the patient to visit technicians for the placement and removal of electrodes; they are costly to health systems; they cannot be given to patients indefinitely; and they require patients to take the initial step of visiting a physician []. This is particularly important, as the asymptomatic patient unaware of their arrhythmia will not present until serious sequelae (eg, stroke secondary to AF) occur. Furthermore, patients are often monitored for 24-48 hours, which has been shown to miss up to 30% of clinically significant arrhythmias [].
Undoubtedly, consumer-owned smart technology negates many of these limitations. The question of efficacy remains. One of the largest trials to date has been the Apple Heart Study, including detailed data for over 400 patients [,]. In this study, of the 400,000 initially recruited patients, over 2000 (0.5%) received a notification for irregular heart rate. Among patients with detailed data available, the positive predictive value was 0.84 (95% CI 0.76-0.92) for an irregular pulse notification detecting AF. Most studies are restricted to screening for AF, and a systematic review has observed overall sensitivities of around 94% and specificities of 93%-96%, depending on whether a smartphone or smartwatch was used [].
Not only has the physical hardware become more portable and acceptable to patients, but the underlying software interpreting the acquired ECG has also improved drastically over recent years. Automated interpretations from traditional ECG machines have been reported as incorrect between 9% and 35% of interpretations; however, this depends on what rhythm is being evaluated (with AF being a particularly troublesome arrhythmia to diagnose) [,]. Newer smart-device AI can learn and adapt when exposed to a new “learning set” of patient results. By providing vast training sets of data to these algorithms in testing, their overall efficacy is improved, compared with traditional ECG auto interpretation, which relies on applying strict measurement parameters to the ECG presented, without the capacity for learning []. For instance, in one of the seminal papers to describe this breakthrough, a learning set of 109 patients with AF was used, which resulted in the algorithm adjusting its weighting for P-wave absence [,]. This optimized algorithm had a sensitivity of 100% and a sensitivity of 96% compared with the initial values of 87% and 97%, respectively []. In an era of greater connectivity, the potential for crowdsourcing enormous datasets has resulted in more accurate and reliable algorithms, with several proprietary and open-source AF-detection algorithms available currently [,]. This demonstrates how deep learning that can now be used in real time for ECG analyses has the potential to far surpass previous automatic ECG interpretations.
Wearable ECG Monitoring in Clinical Practice
The main use of these devices in clinical practice is the detection or exclusion of arrhythmias. KardiaPro has been approved in the United States for the screening and detection of AF, but has been studied in various other conditions including ventricular dysrhythmias, atrioventricular node re-entrant tachycardia, myocardial ischemia, and electrolyte disturbances [,-]. AF is one of the most investigated applications as it is commonly asymptomatic, has a high prevalence (up to 1.4% of all patients aged >65 years), and can lead to devasting consequences such as stroke and death []. Studies examining the use of wearable ECG technology for screening of AF are broadly supportive; the SEARCH-AF Study used wearable ECG screening in pharmacies and found newly diagnosed AF in 15 patients (1.5%), with an overall prevalence of 6.7% []. A subsequent hypothetical community screening economic analysis extrapolated these results into a cost-effectiveness ratio of US $4066 per quality-adjusted life year gained, and a cost of US $20,695 for the prevention of 1 stroke []. When compared with the average inpatient costs of stroke (estimated at US $20,396 ± $23,256) plus associated outpatient costs (US $17,081 for the first-year plus US $16,689 for every year after), this represents potentially an enormous cost saving [,]. An Australian study using similar technology introduced nurse-led smartphone-based AF screening to general practices. The sensitivity and specificity of the automated algorithm were 95% (95% CI 83%‐99%) and 99% (95% CI 98% ‐100%), respectively, and a new diagnosis of AF occurred in 0.8% of patients []. The evidence base for using these devices in screening at-risk populations is steadily increasing, and several further trials are planned for examining wearable ECG technology in other populations, including children [,,]. Case reports exist of wearable ECG technology detecting cardiac ischemia [] exercise-related arrhythmias in athletes [], and polymorphic ventricular tachycardia [], although these are not as commonly studied as the use of ECG for AF screening.
The reasons for these potential benefits over existing methodologies of AF screening and diagnosis have already been discussed; some of the biggest advantages are that patients are more likely to wear these comfortable, easily accessible devices, faster ECG analysis using AI algorithms with increasing diagnostic accuracy, and that data can be read in real time by physicians. There is also a health service economic incentive, as these devices can be bought by patients themselves for a fraction of the cost of a Holter monitor, at no cost to health systems and comparable efficacy for some dysrhythmias []. Patients themselves are also enthusiastic; a survey of 88 people showed that 82% found the device useful and the use of the device prompted a doctor’s visit in 25% of patients []. While this obviously has a benefit if those patients did have arrhythmia, it does lead to questions surrounding resource use. This leads us to consider the potential limitations of this new technology.
Limitations
This technology is not without its potential drawbacks to both the patient and the clinician. One of the largest technical drawbacks of this technology is its reliance using Lead I. Having only 1 positive and 1 negative electrode will only ever be able to provide a 1-lead view as the potential difference cannot be measured at further points (and thus obtain more leads) without more physical electrodes. It is not even possible to obtain augmented limb leads (which are unipolar and so could practically be created using only 1 positive electrode) as the neutral central terminal (Wilson’s Central Terminal) is created by the average of Lead I, Lead II, and Lead III (3 leads). This can make the interpretation of dysrhythmias more difficult. For instance, having only 1 lead makes diagnosis of conduction delays like a right bundle branch block difficult as the characteristic pattern (rSR’ in V1) is not necessarily visible in Lead I. Having only 1 lead on an extremity also increases the risk of artifacts; without other leads to compare with, artifactual “noise” is more difficult to exclude, and this noise can be amplified by having only 1 loosely attached electrode compared with traditionally several firmly attached electrodes.
One method of circumventing these limitations, however, is by changing the positioning of the positive terminal of the electrode (
Figure 3). By keeping the negative terminal in the right hand and moving the positive terminal to the left leg, the potential difference being measured is in line with Lead II, providing now a 2-lead view of the heart. This has been shown to improve the diagnostic accuracy of some cardiac arrhythmias, especially atrial flutter, which may be more visible in inferior leads []. By simply moving this electrode, the sensitivity for atrial flutter increased from 27.3% to 72.7% [].
There are other patient limitations. Using home ECG monitoring relies on patient technical skill set, as well as financial security to purchase one of these devices, and have consistent internet connectivity. With an aging population, the population that may benefit the most from the detection of occult arrhythmias (ie, older population) may be the group that struggles the most with adopting this technology. In addition, financial cost and consistent internet connectivity may also prove challenges for widespread adaptation.
The other major limitation is the practicality of physician access. Ironically, one of the greatest strengths of these devices (24-hour continuous monitoring for as long as the patient wants) can also be a weakness. Whilst a patient who has this technology now can record an ECG at any point in the day (or night), that does not necessarily mean that they will have timely access to a physician across the same hours. Patients who detect a possible arrhythmia outside of their doctor’s availability may be left with 2 options: wait until an appointment becomes available, worrying all the while about potential strokes or cardiac events; or visit their nearest emergency department. From a resource use standpoint, this becomes worrisome, as in some studies, up to 7.3% of normal ECGs were reported as abnormal (sensitivity 97.1%, specificity 78.5%). Applied to the real world, that means 7 of every 100 normal ECGs may be reported as abnormal, resulting in 7 potentially unnecessary hospital visits per 100 normal ECGs. The question of what to do with patients who present with an abnormal ECG taken on a single lead private device is a vexing one. One potential solution could be rotating on-call physicians to review ECGs as they come through (as these can be sent in real time). However, this will leave open questions of compensation for the physician, and the eternal question raised above: how confident can a physician be based of a 1-lead ECG that there is no further pathology to exclude? What are the medicolegal implications of not fully working up a patient with a single positive trace who then has a devastating cardiovascular event? These issues need to be considered for the clinician to provide safe and sound medical treatment and advice to patients and as the prevalence of these devices rises, these are issues that will be faced by more and more clinicians.
Risk stratification may be useful here. The RITMO study examined whether having a higher screening threshold in elderly patients with hypertension and heart failure would increase AF capture rates. In this study, by stratifying by the stroke risk analysis algorithm, the rates of AF capture increased from the reported 3% at baseline to 13.2% []. By building risk stratification software into these devices, appropriate health care use could perhaps be improved.
Conversely, the lack of follow-up may be another limitation. Institution-provided monitors (eg, Holter monitors) have their data reviewed by physicians, and patient follow-up is initiated in the event of significant dysrhythmias. With consumer-owned devices, there is no assurance of follow-up, even if a significant arrhythmia is detected and the patient alerted. This has been borne out in real-life data, with only 57% of patients in the Apple Health Study with an irregular heart beat notification contacting healthcare providers [].
Conclusions
With an ever-growing health technology sector, wearable biometrics are more and more likely to appear outside of clinical research and into clinical practice. Although the machine taking the recordings becomes smaller and the software interpreting the readings becomes smarter, the underlying principles remain the same as what Einthoven first noticed some 100 years ago. If a clinician is then to have an informed discussion with a patient regarding the use of a wearable ECG device, then they must have confidence in their basic sciences to explain the mechanisms and potential limitations of such a device. With the anticipated explosion of these devices in people’s private lives, questions surrounding this are almost a given, and thus, all clinicians should be well acquainted with the basic sciences of electrocardiography.
Wearable ECG devices have many advantages over existing methods of trace acquisition, but also many potential drawbacks. The ease of use, patient-centered care, and increased availability of ECG monitoring must be balanced with a physician’s duty of care and the potential for false-positive results, creating unnecessary unease and overtesting, as well as technical limitations of the devices themselves. Additional research and guidelines regarding the placement of a potential Lead II view, as well as thorough guidelines regarding data management, confidentiality, and physician workload need to be developed quickly before this technology becomes the standard.
Acknowledgments
We thank Shutterstock illustrator asia11m for the use of the graphics (provided under license).
Conflicts of Interest
None declared.
References
- Kligfield P. The centennial of the Einthoven electrocardiogram. J Electrocardiol. 2002;35 Suppl:123-129. [CrossRef] [Medline]
- Kligfield P, Gettes LS, Bailey JJ, et al. Recommendations for the standardization and interpretation of the electrocardiogram: part I: The electrocardiogram and its technology: a scientific statement from the American Heart Association Electrocardiography and Arrhythmias Committee, Council on Clinical Cardiology; the American College of Cardiology Foundation; and the Heart Rhythm Society: endorsed by the International Society for Computerized Electrocardiology. Circulation. Mar 13, 2007;115(10):1306-1324. [CrossRef] [Medline]
- Rachim VP, Chung WY. Wearable noncontact armband for mobile ECG monitoring system. IEEE Trans Biomed Circuits Syst. Dec 2016;10(6):1112-1118. [CrossRef] [Medline]
- Lubitz SA, Faranesh AZ, Selvaggi C, et al. Detection of atrial fibrillation in a large population using wearable devices: the fitbit heart study. Circulation. Nov 8, 2022;146(19):1415-1424. [CrossRef] [Medline]
- Perez MV, Mahaffey KW, Hedlin H, et al. Large-scale assessment of a smartwatch to identify atrial fibrillation. N Engl J Med. Nov 14, 2019;381(20):1909-1917. [CrossRef] [Medline]
- Zenzes M, Seba P, Portocarrero Vivero-Fäh B. The electrocardiogram on the wrist: a frightening experience to the untrained consumer: a case report. J Med Case Rep. Mar 5, 2023;17(1):79. [CrossRef] [Medline]
- Galvani L. Aloysii galvani de viribus electricitatis in motu musculari commentarius. In: De Bononiensi Scientiarum et Artium Instituto Atque Academia. 1791:363-418.
- Raju TN. The Nobel Chronicles. 1924: Willem Einthoven (1860-1927). Lancet. Aug 1998;352(9139). [CrossRef] [Medline]
- Barold SS. Willem Einthoven and the birth of clinical electrocardiography a hundred years ago. Card Electrophysiol Rev. Jan 2003;7(1):99-104. [CrossRef] [Medline]
- Fisch C. Centennial of the string galvanometer and the electrocardiogram. J Am Coll Cardiol. Nov 15, 2000;36(6):1737-1745. [CrossRef] [Medline]
- Klabunde RE. Cardiac electrophysiology: normal and ischemic ionic currents and the ECG. Adv Physiol Educ. Mar 1, 2017;41(1):29-37. [CrossRef] [Medline]
- Gulizia MM, Casolo G, Zuin G, et al. ANMCO/AIIC/SIT Consensus Information Document: definition, precision, and suitability of electrocardiographic signals of electrocardiographs, ergometry, Holter electrocardiogram, telemetry, and bedside monitoring systems. Eur Heart J Suppl. May 2017;19(Suppl D):D190-D211. [CrossRef] [Medline]
- A R cardiac physiology: membrane potential, action potential, automaticity and electrical vectors. Clinical ECG interpretation. 2019. URL: https://ecgwaves.com/cardiac-electrophysiology-ecg-action-potential-automaticity-vector [Accessed 2019-05-09]
- Dupre A, Vincent S, Iaizzo PA. Basic ECG theory, recordings, and interpretation. In: Iaizzo PA, editor. Handbook of Cardiac Anatomy, Physiology, and Devices. Humana Press; 2005:191-201.
- Becker DE. Fundamentals of electrocardiography interpretation. Anesth Prog. 2006;53(2):53-63. [CrossRef] [Medline]
- Neubeck L, Lowres N, Benjamin EJ, Freedman SB, Coorey G, Redfern J. The mobile revolution--using smartphone apps to prevent cardiovascular disease. Nat Rev Cardiol. Jun 2015;12(6):350-360. [CrossRef] [Medline]
- Inc A. Taking an ECG with the ECG app on Apple Watch Series 4. 2019. URL: https://support.apple.com/en-us/HT208955 [Accessed 2019-05-08]
- Lau JK, Lowres N, Neubeck L, et al. iPhone ECG application for community screening to detect silent atrial fibrillation: a novel technology to prevent stroke. Int J Cardiol. Apr 30, 2013;165(1):193-194. [CrossRef] [Medline]
- Smith WM, Riddell F, Madon M, Gleva MJ. Comparison of diagnostic value using a small, single channel, P-wave centric sternal ECG monitoring patch with a standard 3-lead Holter system over 24 hours. Am Heart J. Mar 2017;185(1097-6744):67-73. [CrossRef] [Medline]
- HOLTER NJ. New method for heart studies. Science. Oct 20, 1961;134(3486):1214-1220. [CrossRef] [Medline]
- Lim YL, Mond H, Michael R, et al. Seven-day holter monitoring detects more significant arrhythmias than 24-hour and 3-day monitoring. Eur Heart J. Jan 25, 2023;44(Supplement_1). [CrossRef]
- Prasitlumkum N, Cheungpasitporn W, Chokesuwattanaskul A, et al. Diagnostic accuracy of smart gadgets/wearable devices in detecting atrial fibrillation: A systematic review and meta-analysis. Arch Cardiovasc Dis. Jan 2021;114(1):4-16. [CrossRef] [Medline]
- Anh D, Krishnan S, Bogun F. Accuracy of electrocardiogram interpretation by cardiologists in the setting of incorrect computer analysis. J Electrocardiol. Jul 2006;39(3):343-345. [CrossRef] [Medline]
- Smulyan H. The computerized ECG: friend and foe. Am J Med. Feb 2019;132(2):153-160. [CrossRef] [Medline]
- Bahrami Rad A, Kirsch M, Li Q, et al. A crowdsourced AI framework for atrial fibrillation detection in Apple watch and Kardia mobile ECGs. Sensors (Basel). Sep 2, 2024;24(17):5708. [CrossRef] [Medline]
- Barbagelata A, Bethea CF, Severance HW, et al. Smartphone ECG for evaluation of ST-segment elevation myocardial infarction (STEMI): Design of the ST LEUIS International Multicenter Study. J Electrocardiol. 2018;51(2):260-264. [CrossRef] [Medline]
- Orchard J, Lowres N, Freedman SB, et al. Screening for atrial fibrillation during influenza vaccinations by primary care nurses using A smartphone electrocardiograph (iECG): A feasibility study. Eur J Prev Cardiol. Oct 2016;23(2 suppl):13-20. [CrossRef] [Medline]
- Tabing A, Harrell TE, Romero S, Francisco G. Supraventricular tachycardia diagnosed by smartphone ECG. BMJ Case Rep. Sep 11, 2017;2017:bcr2016217197. [CrossRef] [Medline]
- Yasin OZ, Attia Z, Dillon JJ, et al. Noninvasive blood potassium measurement using signal-processed, single-lead ecg acquired from a handheld smartphone. J Electrocardiol. 2017;50(5):620-625. [CrossRef] [Medline]
- Freedman B, Camm J, Calkins H, et al. Screening for atrial fibrillation: a report of the AF-SCREEN international collaboration. Circulation. May 9, 2017;135(19):1851-1867. [CrossRef] [Medline]
- Lowres N, Neubeck L, Salkeld G, et al. Feasibility and cost-effectiveness of stroke prevention through community screening for atrial fibrillation using iPhone ECG in pharmacies. The SEARCH-AF study. Thromb Haemost. Jun 2014;111(6):1167-1176. [CrossRef] [Medline]
- Godwin KM, Wasserman J, Ostwald SK. Cost associated with stroke: outpatient rehabilitative services and medication. Top Stroke Rehabil. Oct 2011;18 Suppl 1(sup1):676-684. [CrossRef] [Medline]
- Wang G, Zhang Z, Ayala C, Dunet DO, Fang J, George MG. Costs of hospitalization for stroke patients aged 18-64 years in the United States. J Stroke Cerebrovasc Dis. 2014;23(5):861-868. [CrossRef] [Medline]
- Gwynne K, Flaskas Y, O’Brien C, et al. Opportunistic screening to detect atrial fibrillation in Aboriginal adults in Australia. BMJ Open. Nov 15, 2016;6(11):e013576. [CrossRef] [Medline]
- Gropler MRF, Dalal AS, Van Hare GF, Silva JNA. Can smartphone wireless ECGs be used to accurately assess ECG intervals in pediatrics? A comparison of mobile health monitoring to standard 12-lead ECG. PLoS One. 2018;13(9):e0204403. [CrossRef] [Medline]
- Provencio A, Gil MÁC. Smartwatch electrocardiogram records ST depression, premature ventricular complexes, and ventricular fibrillation. Lancet. Nov 12, 2022;400(10364):e12. [CrossRef] [Medline]
- Sanchis-Gomar F, Lavie CJ, Perez MV. Consumer wearable technologies to identify and monitor exercise-related arrhythmias in athletes. Curr Opin Cardiol. Jan 2021;36(1):10-16. [CrossRef] [Medline]
- Avidan Y, Tabachnikov V, Danon A, Schliamser JE. Polymorphic ventricular tachycardia detected by a smartwatch in a patient with recurrent syncope. JACC Case Rep. Oct 2, 2024;29(19):102606. [CrossRef] [Medline]
- Rajakariar K, Koshy AN, Sajeev JK, Nair S, Roberts L, Teh AW. Modified positioning of a smartphone based single-lead electrocardiogram device improves detection of atrial flutter. J Electrocardiol. 2018;51(5):884-888. [CrossRef] [Medline]
- Andrade RD, Vitorino PVO, Sousa ALL, et al. A program to optimize the detection of paroxysmal atrial fibrillation: the RITMO study. Arq Bras Cardiol. 2024;121(9):e20240235. [CrossRef] [Medline]
Abbreviations
| AF: atrial fibrillation |
| AI: artificial intelligence |
| ECG: electrocardiogram |
Edited by Andrew Coristine; submitted 29.05.24; peer-reviewed by Irena Jekova, Ivo Iliev; final revised version received 22.12.24; accepted 23.12.24; published 10.02.25.
Copyright© Samuel Smith, Shalisa Maisrikrod. Originally published in JMIR Cardio (https://cardio.jmir.org), 10.2.2025.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Cardio, is properly cited. The complete bibliographic information, a link to the original publication on https://cardio.jmir.org, as well as this copyright and license information must be included.