Electrocardiographic artifacts in clinical practice: A logical approach to recognition and prevention

Abstract

Electrocardiographic (ECG) artifacts are anomalous signals unrelated to cardiac electrical activity that interfere with our interpretation and management. They can mimic serious arrhythmias or ischemic changes and often lead to unnecessary interventions. Understanding why such artifacts occur requires a grasp of how ECG signals are recorded, from tiny millivolt-level cardiac signals traveling from the heart through skin electrodes, cables, amplifiers, filters, and finally the display on the monitors or ECG strip. At each step, artifacts may enter, whether from patient movement, tremor, electrode-skin interface problems, broken leads, environmental interference, or implanted devices. Although filters and digital processing aim to reduce noise, artifacts may still occur due to the significant overlap of the noise frequencies with those of the true cardiac signals, or they can arise from non-electrical causes. Recognition of these artifacts requires some logic: Correlating the ECG with patient status, assessing physiological plausibility, checking lead distribution, and confirming with ancillary signals. This review summarizes the physics of ECG recording in simple terms, classifies common artifacts and outlines practical bedside clues and preventive strategies. The goal is to provide a clear, clinically oriented framework so that artifacts are recognized not as rare technical nuisances but as everyday challenges that can be safely navigated through reasoning and awareness.

Key Words: Electrocardiographic; Electrocardiogram; Artifacts; Pulse tapping; Square wave; Electromagnetic interference; Orphan

Core Tip: Electrocardiographic artifacts are non-cardiac signals that corrupt monitoring, often mimicking serious arrhythmias or ischemia and leading to unnecessary interventions. They are introduced at various points in the recording chain - from patient movement and electrode issues to environmental interference - due to the tiny cardiac signal and frequency overlap with noise. Recognizing these artifacts is vital; it requires clinical correlation, checking physiological plausibility, and assessing lead distribution to avoid misdiagnosis and safely manage the patient. This review summarizes the physics of electrocardiographic recording in simple terms, classifies common artifacts and outlines practical bedside clues and preventive strategies.

INTRODUCTION

Electrocardiograms constitute the cornerstone of cardiovascular diagnosis. In modern clinical practice, electrocardiographic (ECG) interpretation guides decisions in the emergency department, intensive care unit (ICU), operating theatre, and outpatient clinic and, increasingly, in patients’ homes through wearable devices. Its advantages include technical simplicity, cost-effectiveness, and the ability to deliver instant information about the heart’s rhythm and ischemic status. However, its greatest weakness lies in its fragility: The signal captured at the skin surface is relatively small and vulnerable to distortion for various reasons, resulting in artifacts[1-4].

Artifacts, which are signals not generated by the heart, are common in daily practice and are often overlooked. They can be obvious and inconsequential at best and misleading and capable of mimicking life-threatening arrhythmias, including ventricular tachycardia (VT) or fibrillation, as well as ischemic ST-T changes at worst[1-4]. Unrecognized artifacts can trigger cascades of unnecessary interventions, such as antiarrhythmic infusions, serum drug concentration testing, suspecting drug toxicity, inappropriate defibrillation, catheterization or even intensive care admission[5-9]. Case reports and series repeatedly show that patients subjected to therapies for arrhythmias were later proven to be artifacts[5].

Despite their frequency and clinical consequences, artifacts are not routinely emphasized in medical training. As a result, it is common for physicians to recognize artifacts in a pattern-recognizing manner without understanding the underlying technical nuances. Without proper logic, recognition becomes inconsistent and heavily reliant on experience. This review aims to provide insights into the technical aspects of signal recording, the causes of common issues and ways to prevent them, and the troubleshooting when they persist. Here, we present a classification of artifacts, explore their most important clinical mimics and outline practical recognition strategies. Finally, we discuss prevention and emerging solutions, including artificial intelligence (AI).

PHYSICS OF ECG RECORDING

The fragile nature of the ECG signal

The ECG records very small voltages, typically in the range of 0.5-5 mV. For comparison, a standard household light bulb operates at 220 Volts in India or 110 Volts in the United States. The ECG signal is therefore a million times smaller and can be easily overshadowed by competing signals from skeletal muscle or electrical interference in the environment. The body itself is electrically noisy. Skeletal muscle contraction produces signals that are up to 10 times greater than cardiac potentials. Respiration alters thoracic impedance, causing slow drifts at baseline. Even the act of speaking introduces movement that can distort signals.

The signaling pathway

To understand artifacts, tracing the journey of the ECG signal is essential (Figure 1, Table 1).

Figure 1 Journey of an electrocardiogram (from signal generation to printed output). Possible source of artifacts in each step.

Table 1 Signal pathway and artifact risks.

Step	Function	Common artifact sources	Clinical example
Generation	The heart produces approximately 1 mV currents	Competing signals from the muscle	Myopotential artifacts mimicking atrial fibrillation
Detection	Electrodes capture surface signals	Sweat, hair, poor adhesion, movement	Wandering baseline or false ST elevation (baseline drift)
Transmission	Leads carry signals	Cable fracture, patient motion, EMI	Flatline due to broken lead; saw-tooth interference from cautery
Amplification and filtering	Boosts and cleans signals	Amplifies both signal and noise; filter distortion	QRS smoothed, or ST segments flattened, spurious ST shifts
Display and printing	Trace presented	All upstream noise is now visible	Artifact interpreted as ventricular tachycardia, acceleration or deceleration artifacts

EMI: Electromagnetic interference.

Generation: Myocardial depolarization generates action potentials that summate into measurable electrical fields.

Detection: Surface electrodes act as receivers, picking up these potentials. Their performance depends on skin preparation, electrode quality, and patient factors that affect good contact of the electrode with the skin (sweat, movement, tremor).

Transmission: Leads and cables transmit the signal to the ECG machine. Mechanical motion, cable fracture, or electromagnetic interference can corrupt transmission.

Amplification and filtering: The amplifiers enhance the small signal but also amplify unwanted noise. In contrast, the filtering removes specific frequencies, thus retaining only the spectrum of frequencies containing the cardiac signals. However, filtering is not perfect, as some of the noises overlap with the frequency spectrum of the signals.

Display and printing: The final waveform is shown on screen or paper. By this point, artifacts are indistinguishable from true cardiac activity unless identified through logic. The final printing is subsequently performed on calibrated ECG paper.

Signal processing and filters

Amplification - the indiscriminate booster: Amplifiers are crucial for making tiny ECG signals visible. However, they are nonselective and amplify every signal they receive, whether it is a true QRS complex, skeletal muscle tremor, or electrical hum from nearby equipment[1]. Analogy: An amplifier is similar to a microphone in a crowded hall. It makes the whisper of the heart audible, but it also amplifies coughs, chair scrapes, and background chatter.

Filters - sieves of the ECG: Filters aim to retain useful cardiac signals and remove unwanted noise by filtering off the frequencies of particular bandwidths and are displayed as footnotes in ECGs (Figure 2, arrow).

Figure 2 All the electrocardiograms have a notch filter (60 Hz in this example) and the low- to high-pass filter mentioned in the footnote (arrow).

High-pass filters (cutoff approximately 0.05-0.5 Hz): Allow any signal of higher frequency than their cutoff value. Remove very slow drifts, such as baseline wander caused by respiration or electrode movement.

Low-pass filters (cutoff approximately 100-150 Hz for diagnostic ECGs): Allow frequencies lower than their cutoff and remove signals of higher frequency, including tremors or electrical buzzes.

Notch filters (50 Hz in India, 60 Hz in the United States): Remove specific frequency interference from the power supply lines.

Bandpass filters (0.05-150 Hz): Allow only the typical frequency range of cardiac signals. Analogy: Filters work like sieves. They can block very large particles (slow baseline wander) or very fine dust (high-frequency tremors). However, when unwanted particles are the same size as the intended grains, separation becomes impossible.

Why filters cannot eliminate all artifacts

Despite sophisticated filtering, artifacts persist (Figure 3). The reasons include the following.

Figure 3 Frequencies of cardiac signals and artifacts showing significant overlap. P and T waves overlap with tremors; ST shifts overlap with baseline wander; QRS overlaps with 50 Hz interference. ECG: Electrocardiogram.

Frequency overlap: Tremor (4-10 Hz) overlaps with P- and T-wave frequencies. Removing tremors risks removing true signals.

Risk of distortion: Excessive filtering can smooth sharp deflections, flatten ST segments, or widen QRS complexes. This may create false pathology.

Unpredictable noise: Motion artifacts are irregular and cannot be predicted or filtered, such as steady 50 Hz interference.

Non-electrical faults: Loose electrodes or cable breaks produce sudden baseline shifts. These are not frequency-based or bypass filters.

No clinical reasoning: Machines cannot judge physiologic plausibility. A monitor may display “ventricular fibrillation” while the patient is talking comfortably.

Practical clinical implications

Filters can lessen but not completely eliminate artifacts in ECG recordings, and excessive filtering may also obscure important diagnostic information by distorting true waveforms. Modern ECG devices often provide both filtered and unfiltered traces, enabling clinicians to compare and better differentiate artifacts from genuine signals. Ultimately, the ECG should never be interpreted in isolation, and correlation with the patient’s clinical context remains essential.

CLASSIFICATION OF ARTIFACTS

Artifacts can be classified in several ways, but for practical bedside teaching, a clinical classification is most effective. We categorize them into four main groups: Patient-related, equipment-related, environmental, and device-related (Table 2).

Table 2 Classification of electrocardiographic artifacts with examples.

Category	Source	ECG appearance	Clinical mimic	Solution
Patient-related	Tremor, shivering, respiration, movement	Irregular baseline, wandering drift	AF, flutter, VT, ST changes	Patient cooperation
Equipment-related	Loose electrodes, fractured cables, faulty machine	Baseline jumps, flatline, spikes	Asystole, VF	Change cables, better skin preparation, electrode contact
Environmental	Power-line hum, cautery, ventilators	Saw-tooth oscillations, rhythmic noise	Flutter, VT	Avoidance of equipment, notch filters
Device-related	Pacemaker, LVAD, neurostimulators, IVL	Sharp spikes, continuous noise	VF, tachycardia	Not always possible

ECG: Electrocardiogram; AF: Atrial fibrillation; VT: Ventricular tachycardia; VF: Ventricular fibrillation; LVAD: Left ventricular assist device; IVL: Intravascular lithotripsy.

Patient-related artifacts

These are the most common sources of spurious ECG signals caused by the patient’s physiological or mechanical movements. Skeletal muscle potentials, which are much greater than cardiac potentials, are easily recorded when muscles contract near the electrodes. This causes quick, irregular baseline oscillations that can resemble atrial fibrillation or coarse undulations from shivering that might be mistaken for atrial flutter or VT (Figure 4A). Gross body movements can also disrupt electrode contact or stretch cables, producing large deflections that may mimic wide-complex tachycardia, sometimes confused as “polymorphic VT” until proven otherwise (Figure 4B). Even breathing affects thoracic impedance and electrode position, leading to slow baseline drift that can resemble ST elevation or depression (Figure 4C). A key point is that when the patient moves, shivers, or trembles, the clinician should strongly suspect artifacts, and interpretation must be based on the clinical context (Figure 4D). Respiration may also cause low-frequency artifacts in ECGs as well as in Holter tracings (Figure 5A and B). Mechanical movements usually result in low-frequency artifacts and can mimic flutter or atrial fibrillation. In contrast, myopotential artifacts, as observed in fine tremors, fasciculations, etc., cause high-frequency artifacts that mimic attenuation correction artifacts. However, myopotential artifacts result in irregular high-frequency distortions, whereas the attenuation correction artifacts are regular and have the same frequency as the alternating current (50 Hz or 60 Hz) (Figure 5C).

Figure 4 Patient-related motion artifacts. A: Classical Parkinson’s rest tremor. Note that the frequency of the artifact is approximately 5 Hz (200 microseconds interval). Note that lead 2 does not have an artifact, indicating that the artifact is coming from the left arm; B: Limb movement artifacts resembling polymorphic ventricular tachycardia. Note that the normal QRS complexes (orange circles) are marching through the artifacts; C: Intraprocedural spurious ST elevation due to lead movement (attached to the subcostal region) during respiration; D: Trembling during micturition, causing baseline high-frequency irregular artifacts; note that the stoppage of the act of micturition resolves the baseline artifact.

Figure 5 Patient-related and equipment-related artifacts. A: Low-frequency artifacts (0.7 Hz) can be seen marching through without any PQRST association (orange circles). While the source of the artifact is unclear, they are too slow to be physiological tremors, which fall in the range of 3-12 Hz (or even higher). Respiratory movements can typically cause artifacts in this range (0.7 Hz), but the odd point corresponds to a respiratory rate of approximately 42/minute. The heart rate is too slow for that type of tachypnea; B: Similar artifacts can be seen on a Holter tracing from a young child, likely a respiration artifact (arrows); C: Attenuation correction artifacts showing low-amplitude high-frequency noise picked up by V2 lead to a crying newborn (D/D intercostal muscle fasciculation); D: Poor electrode contact causes intermittent baseline jumps in V4 and V5. Notably, artifacts may coexist with life-threatening arrhythmias. In this example, the patient had bradycardia-related Torsades de Pointes with simultaneous artifacts, emphasizing clinical evaluation while interpreting or suspecting artifacts on electrocardiogram; E: Poor electrode contact causes irregular baseline jumps in V6 (arrows).

Equipment-related artifacts

Another key source of interference comes from faults in electrodes, cables, or the ECG machine itself. Poor electrode contact, often caused by dried gel, sweat, or insufficient adhesion, weakens electrical connectivity and results in intermittent baseline jumps or flatlines in the affected leads (Figure 5D and E). Likewise, fractured lead wires or loose connectors may cause sudden spikes, dropouts, or continuous signal loss; a common example includes faulty lead wires producing tracings that look indistinguishable from ventricular fibrillation. Machine-related problems are currently less common but still relevant: Irregular paper speed (Figure 6A-C) in older analog systems, excessive filter settings that flatten/elevate ST segments (Figure 6D and E), or amplifier saturation that causes clipped waveforms. Sometimes, an ECG analog may fail to process part of the electrical activity when a segment of a PQRST complex is missing (Figure 7A).

Figure 6 Equipment-related artifacts. A: Deceleration artifact (orange box) causing narrowing of the QRS complex and giving a false impression of a spurious atrial premature contraction (arrow) advancing the next QRS complex (note the further narrowing of the QRS complex due to paper deceleration) in an ongoing narrow complex tachycardia; B: Deceleration artifact (orange box) causing narrowing of the QRS complex (nearly a straight vertical line - arrows) (note the narrowing of the preceding and the next T waves due to paper deceleration); C: Acceleration artifacts cause stretching of the ST segment, resulting in a false impression of sinus slowing. Prolonged stretching may lead to false impressions of significant sinus pauses or atrioventricular blocks. Note that the initial paper deceleration caused narrowing of the QRS complex, followed by paper acceleration; D: Filtering error causing a sharp artifact after the QRS complex, resulting in a false impression of double-QRS (arrows); E: Similar error in filtering causing significant distortion of waveforms with a baseline shift, making the electrocardiogram incomprehensible.

Figure 7 Equipment-related and environmental artifacts. A: Missing PQRSs due to processing failure with residual T waves give rise to double T waves or “orphan T” waves (arrows); B: Arterial pulse tapping artifact mimicking ST elevation. All QRS complexes are followed by artifacts with electromechanical associations. Note that lead 2 is spared, meaning that the artifact comes from the left arm electrode; C: Similarly, arterial pulse tapping artifacts mimic bizarre T waves with prolonged QT intervals. Note that lead 1 is spared, meaning that the artifact comes from the left lower limb electrode; D: Electrocardiogram showing fixed-frequency square wave artifacts. ECG: Electrocardiogram.

Baseline drift is a low-frequency artifact where the isoelectric line on an electrocardiogram tracing slowly moves up and down instead of remaining flat. It is one of the most common artifacts encountered and is usually due to poor contact between the skin and the electrode. This can be mitigated by improving contact, better skin preparation, the use of appropriate high-pass filters, or more sophisticated signal processing techniques such as wavelet transform and independent component analysis[10,11].

The arterial pulse-tap artifact, also known as Aslanger’s sign or electromechanical association artifact, is caused by an electrode being placed over a pulsating artery (e.g., radial, brachial, axillary, or even the apical impulse on the chest, an arteriovenous fistula created for hemodialysis)[12,13]. The mechanical vibrations of the arterial pulsation are picked up by the electrode and superimposed on the electrical signal of the heart. The key characteristic is that the artifact is regular and occurs every cardiac cycle, usually at a constant time delay after the QRS complex, corresponding to the pulse transit time to that location, mimicking an ST elevation or prolonged QT interval (Figure 7B and C). A repeat ECG with repositioning of the electrode easily removes the artifact.

Environmental artifacts

The hospital environment is filled with electrical devices that can cause interference in ECG recordings. Power-line interference is the most common type of interference, occurring as regular sawtooth waves across all leads, corresponding to the local alternating current frequency (50 Hz in India and 60 Hz in the United States). Interestingly, wires connected from the patient’s body to cardiac monitors may produce square wave artifacts (Figure 7D)[3]. In operating theatres and ICUs, electrocautery produces high-frequency interference that may resemble flutter waves or VT (Figure 8A), whereas infusion pumps, ventilators, and dialysis machines generate rhythmic oscillations aligned with device activity. Even static electricity from blankets, clothing, or staff-handling electrodes can cause sharp baseline jumps, making interpretation more difficult.

Figure 8 Environmental and device-related artifacts. A: Electromagnetic interference causes sharp deflections, mimicking monomorphic ventricular tachycardia; B: Holter strip showing pseudopacing artifacts that cannot be explained by any device algorithm; C: Electrocardiogram in a patient with a dual-chamber pacemaker showing pseudopacing artifacts (arrows); D: Deep brain stimulators causing high-frequency (usually 60-200 Hz) artifacts (12 pulses in 200 microseconds = 60 Hz) (arrow - zoomed snippet of V2); E: During intravascular lithotripsy, shock pulses can be detected via electrocardiogram. It can even capture the ventricles, which are called shocktopics (arrows). For each cycle of intravascular lithotripsy shock, there are 10 pulses at 1-second intervals (1 Hz).

Device-related artifacts

Finally, modern implanted or external devices have introduced new sources of artifacts. Pacemakers are the most well-known example, with pacing spikes often exaggerated or distorted by filtering, sometimes leading to misinterpretation as ventricular fibrillation. Many pseudopacing artifacts can also be observed, which cannot be explained by any pacing algorithm (Figure 8B and C). Owing to their continuous mechanical rotation, left ventricular assist devices produce persistent high-frequency noise that can completely obscure the underlying cardiac rhythm. Neurostimulators - including deep brain, sacral nerve, and gastric stimulators - generate periodic pulses that may overlap with ECG frequencies, adding another layer of complexity (Figure 8D). During intravascular lithotripsy, shock pulses can be detected via ECG. It can even capture, which is called shocktopics (Figure 8E). Even the watch dial can create an artifact at a 1 Hz frequency corresponding to its second dial. As these devices become more common, clinicians must remain aware of their potential to create misleading ECG patterns.

DISTRIBUTION OF ARTIFACTS ACROSS LEADS

A distinctive but clinically significant aspect of artifact recognition involves understanding why disturbances impact certain leads while sparing others. This is best explained by the role of the Wilson Central Terminal (WCT), which serves as the reference point for all precordial (chest) leads. The WCT is derived by averaging the electrical potentials from the three limb electrodes - right arm, left arm, and left leg. Consequently, when a limb electrode produces noise or becomes unstable, the WCT becomes contaminated, and this disturbance propagates to all chest leads while also distorting the limb leads themselves. Conversely, if a single chest electrode is faulty, the artifact remains confined to that specific lead - for example, noise limited to V3. Since this mechanism is localized to the electrode over the artery, the artifact usually spares one lead completely and affects different leads in a graded manner (Figure 9). Finally, when all leads are affected simultaneously, the source is almost always equipment-related or environmental, such as electrocautery interference or power-line noise. A useful teaching point is that by analyzing the distribution of artifacts across different leads, clinicians can often pinpoint the faulty electrode or differentiate between patient-, equipment-, and environment-related noise.

Figure 9 Approach to artifact detection. Schematic depiction of affected leads when an artifact is generated from a specific limb; color facilitates easy interpretation. LA: Left arm; RA: Right arm; LL: Left leg; Vx: Precordial lead; WCT: Wilson’s central terminal.

CLINICAL MIMICS AND PITFALLS

Artifacts become clinically significant not because they exist but because they closely resemble disease (Table 3). The greatest danger is when an artifact mimics a life-threatening arrhythmia or an ischemic event, potentially leading to misdiagnosis and inappropriate treatment. Several common mimics illustrate the importance of careful recognition.

Table 3 Common clinical mimics of electrocardiographic artifacts.

Artifact source	ECG appearance	Clinical mimic	Recognition clues
Tremor (Parkinson’s, shivering)	Irregular baseline, absent P waves	Atrial fibrillation	P waves marching through; stable pulse
Electrocautery, ventilators	Saw-tooth baseline	Atrial flutter	Localized, no AV conduction ratio
Movement, loose cables	Wide irregular complexes	Ventricular tachycardia	Too fast; QRS visible through noise; stable patient
Loose electrode, EMI	Chaotic baseline	Ventricular fibrillation	Conscious patient; stable SpO2
Respiration, poor skin contact	Wandering baseline, pseudo ST shift	Myocardial infarction	Inconsistent, resolves with repeat
Holter/wearable motion	Drop-outs, spikes	AV block, tachycardia	Correlate with activity; not reproducible

ECG: Electrocardiogram; EMI: Electromagnetic interference; AV: Atrioventricular.

Atrial fibrillation mimics are often caused by tremors or shivering, which introduce irregular, high-frequency oscillations into the baseline. These oscillations obscure P waves and can make the QRS complexes appear irregular, resulting in a tracing that closely resembles atrial fibrillation. However, careful examination may reveal P waves that appear to “march through” the noise, and unlike true fibrillation, tremors usually produce uniform oscillations of the same amplitude. Clinical correlation is essential: A patient with a regular pulse despite an apparently irregular ECG almost certainly experiences an artifact.

Atrial flutter (Figure 4A) mimics often results from rhythmic external oscillations such as electrocautery or ventilators. These produce saw-tooth baseline deflections, especially in the inferior leads, and can be mistaken for flutter waves. In genuine flutter, atrial activity remains consistent across all leads and shows a consistent atrioventricular conduction pattern (for example, 2:1, 3:2 atrioventricular conduction, etc.). Conversely, artifacts tend to be localized and lack physiological conduction ratios. The patient’s hemodynamic stability and specific lead involvement (e.g., Parkinson’s tremor) further help differentiate artifacts from true flutter.

VT (Figures 4B and 8A) mimics are among the most concerning pitfalls. Large, rapid oscillations caused by gross body movement, shivering, or loose lead cables can resemble wide-complex tachycardia. These may appear chaotic or polymorphic, raising concerns of malignant arrhythmia. Clues to identification include the apparent “rate”, which often exceeds 400-600 beats per minute - much faster than true VT, which rarely exceeds 250-300 bpm. Furthermore, the patient’s true rhythm may show faintly visible QRS complexes “marching through” the artifact. Most importantly, patients are often alert and hemodynamically stable, a scenario incompatible with genuine sustained VT. Case reports describe near-interventions caused by such artifacts, highlighting the need for vigilance.

Ventricular fibrillation mimics are another critical danger, as VFs require immediate defibrillation (Figure 4B). Artifacts caused by electrode detachment, tremor, or electrical interference can create a chaotic baseline that looks like fibrillation. Recognizing this involves checking the patient: A conscious, talking patient is not in VF. Simultaneous monitoring of SpO₂ or arterial pressure usually reveals an organized rhythm. Additionally, VF artifacts often spare some leads, unlike true ventricular fibrillation, which affects all leads.

ST-segment elevation can mimic true changes due to baseline drift, pulse tapping of the electrodes, poor electrode-skin contact, or respiratory motion (Figures 4C and 7B). These artifacts cause apparent ST elevation or depression and may be mistaken for acute myocardial infarction. The inconsistency of ST changes between beats and across leads is a key diagnostic clue, whereas genuine infarction tends to produce stable and contiguous lead alterations. Simply repreparing the skin and repeating the ECG can eliminate artifacts and avoid unnecessary interventions such as thrombolysis or catheterization. Pacemaker artifact misinterpretation occurs because pacing spikes are often exaggerated or distorted by filters. Tall or broad spikes may be mistaken for fibrillation or electrical noise. Recognition requires awareness of the pacing mode: Spikes occurring at predictable intervals in relation to atrial or ventricular events are almost always pacing artifacts rather than pathology (Figure 8B and C).

Wearable and Holter monitors add another layer of complexity. Movement, sweat, and poor adhesion often cause baseline drift, false pauses, or spurious tachyarrhythmias. For example, electrode detachment can imitate high-grade atrioventricular block, whereas exercise-related noise may be mistaken for tachycardia. These artifacts can lead to unnecessary referrals for pacemaker implantation if tracings are not reviewed carefully and in context.

Finally, pediatric ECGs are especially prone to artifacts (Figure 5C). Children’s frequent movements, crying, or shivering create tremor artifacts that often shift limb leads. The smaller chest size also makes electrode contact less stable, further complicating interpretation. Extra care and clinical correlation are therefore crucial when assessing pediatric tracings.

RECOGNITION LOGIC FOR CLINICIANS

Recognizing ECG artifacts requires clinical judgment rather than simple pattern detection. Machines and algorithms cannot reliably differentiate artifacts from true arrhythmias, especially when they overlap, so the clinician bears responsibility. A structured, step-by-step method ensures consistency, whether at the bedside, in the ICU, or when reviewing Holter and wearable tracings (Figure 10). The initial and most vital step is to assess the patient before analyzing the ECG. A tracing that seems to indicate ventricular fibrillation might actually be an artifact if the patient is awake, talking, and has a normal pulse and blood pressure. The principle is to assess physiologic plausibility. Human conduction has specific limits: Sinus rates rarely exceed 220 beats per minute in adults, atrioventricular conduction usually does not exceed 250-300 beats per minute, and VT above 300 beats per minute is practically impossible. Apparent ventricular “rates” of 400-600 beats per minute, which are often caused by artifacts, are outside human physiological capabilities and therefore cannot reflect a true rhythm. Bedside correlation with clinical status remains the quickest and most dependable method to avoid disastrous misinterpretation.

Figure 10  A stepwise approach to electrocardiographic artifacts. Orange arrow - spurious atrial premature complex, see Figure 6A; blue arrow - orphan T wave of the corresponding missing QRS complex along with normal T wave, color facilitates easy interpretation. LA: Left arm; RA: Right arm; LL: Left leg; Vx: Precordial lead; WCT: Wilson’s central terminal.

The next step involves carefully examining the baseline, which often reveals artifacts before the QRS complexes do. A wandering baseline suggests respiration or gradual electrode drift. Sudden baseline jumps indicate poor electrode contact or static discharge. A regular saw-tooth oscillation strongly points to power-line interference, whereas coarse, rhythmic oscillations synchronized with patient shivering indicate tremor artifacts. The next step is to identify true P/QRS/T complexes. In an artifact, genuine P/QRS/T complexes often “march through” the noise with regular intervals (irregular in concomitant atrial fibrillation or other irregular atrial arrhythmias), whereas in true arrhythmia, the baseline and QRS morphology change together. In true monomorphic VT, the interval between successive QRS complexes (the R-R interval) is typically regular or nearly regular. In movement artifacts mimicking VT, the “R-R” intervals of artifactual deflections are often grossly irregular. Recognition of underlying QRS complexes has saved many patients from unnecessary interventions; for example, apparent wide-complex tachycardia was later recognized as a tremor artifact once regular QRS complexes were detected beneath the oscillations.

The next step is to evaluate for electromechanical associations. Arterial pulse tapping artifacts, also known as electromechanical association artifacts, are ECG distortions that arise when mechanical pulsations from an underlying artery, such as the radial or posterior tibia, are transmitted to an adjacent limb electrode. This mechanical influence can then create anomalous waveforms on the ECG tracing immediately after each QRS. Any repetitive waveform with a fixed, short coupling interval after the QRS should trigger suspicion for a mechanical, noncardiac source. If no electromechanical association is found, one should assess how the artifact is distributed across leads. If all leads are equally affected, a common cause, such as equipment malfunction or environmental interference, is most likely. If one limb lead is spared, a loose limb electrode is likely responsible, since instability of a single-limb electrode can contaminate the WCT and thus affect all the chest leads and all the limb leads except one. Conversely, artifacts confined to a single chest lead indicate local electrode issues. The pattern of artifacts, therefore, acts as a map to identify the faulty electrode.

Finally, the artifacts can be seen as additional findings in the ECG, as a distortion of the existing and expected wave forms or, rarely, may appear as a deletion of the expected wave forms. The initial step would be timing the suspected artifact with the cardiac cycle in the case of repetitive artifacts. Noting whether the artifact is high-frequency vs low-frequency helps, since the causes differ. Examples of high-frequency artifacts include attenuation correction artifacts, myopotential artifacts, electromagnetic interference, and phantom pacing spikes. Examples of low-frequency artifacts are baseline drift, shaking, and arterial pulse tap artifacts. Artifactual deletion of expected waveforms or pacing spikes is more often observed in analog-type monitors than in usual ECGs. Artifactual distortion of waves is also common in such systems and may affect ST analysis. This is the major reason that ICU monitors are not preferred for ST-segment analysis. In analog systems, artifacts are a more significant challenge because filters are hardware-based and less sophisticated than modern digital signal processing algorithms. Correcting an artifact often requires retaking the entire ECG physically, as post-processing options are limited or nonexistent. Accurate interpretation relies heavily on the skill of the technician in acquiring a clean signal and the clinician in visually distinguishing artifacts from genuine cardiac pathology.

PREVENTION AND TROUBLESHOOTING

Artifacts are not merely nuisances but also preventable sources of error in ECG interpretation. A structured prevention strategy not only saves time and avoids misdiagnosis but also reduces the risk of unnecessary interventions. The principles of prevention can be summarized in four key domains. First, clinicians should anticipate artifacts in settings where they are especially common, such as in shivering patients, in operating theatres with electrocautery, or in ICUs where multiple electrical devices operate simultaneously. Second, optimization of the recording setup is essential, with careful attention given to electrode placement, secure contact, and proper lead management. Third, the recording environment should be controlled by minimizing sources of external interference. Finally, staff education remains crucial, as recognition of artifacts should be a routine part of training for nurses, technicians, and residents.

At the patient level, several strategies are important. Skin preparation should be meticulous: The skin must be clean and dry to reduce impedance, excessive chest hair should be shaved, and electrodes should be placed away from bony prominences or areas prone to sweating. Patient comfort also plays a role; warming a shivering patient, encouraging muscle relaxation, and minimizing talking or movement during a standard 12-lead recording all reduce noise. Proper positioning can further stabilize the tracing - supporting tremulous limbs - and using pediatric-sized electrodes in children improves electrode contact and reliability.

At the equipment level, the focus is on electrodes, cables, and machines. Fresh electrodes should always be used, as dried gel increases impedance and noise, and firm adhesion must be ensured. In the ICU and operating theatre, patients undergoing continuous monitoring, the electrodes should be replaced regularly. Cables and leads should be inspected for cracks, fractures, or stretching and secured to prevent tugging or accidental dislodgement. ECG machines must be maintained with appropriate calibration of filters; overfiltering should be avoided, and “unfiltered” displays should be reviewed when necessary to preserve diagnostic clarity. Electrical grounding and routine service are equally important. Environmental measures also contribute to artifact prevention. ECG cables should be kept separate from power cords and electrocautery units, and unnecessary electrical devices should be switched off during recording. In the operating theatre, tracings are best obtained during pauses in electrocautery, and shielded cables may be employed in environments with high electromagnetic interference.

When artifacts do occur, a simple troubleshooting algorithm provides the following structure: First, check the patient and determine whether they are stable or unstable; second, check the electrodes, reposition them and reprepare the skin if needed; third, inspect leads and cables, replace them if suspicious; fourth, check for environmental interference; fifth, record them after corrections; and sixth, crosscheck with ancillary signals such as SpO₂ or arterial line tracings. Staff education is the cornerstone of prevention. A formalized “prerecording checklist”, encompassing verification of skin preparation, proper electrode placement, cable integrity, and patient immobility, should be implemented to prevent the formation of ECG artifacts, thereby significantly abating the chain of diagnostic tests. Nurses, technicians, and residents must be trained to recognize artifact patterns and apply simple but critical rules: Never shock a talking patient, always check electrode contact before diagnosing arrhythmia, and always correlate the ECG with pulse oximetry or arterial waveform before escalating care.

ADVANCES IN AUTOMATED ARTIFACT DETECTION

Despite significant technological advances, distinguishing artifacts from genuine arrhythmia remains a major challenge for automated systems. While digital filtering and AI provide promising solutions, human clinical judgment remains essential. Digital filtering techniques have advanced significantly. Adaptive filters now adjust automatically to changes in the baseline, reducing interference without the excessive smoothing typical of fixed filters. Signal averaging methods can also diminish repetitive noise patterns, improving the clarity of tracings while maintaining clinically relevant information. These methods improve recording quality but cannot eliminate all sources of artifacts, especially those caused by mechanical motion or electrode misplacement.

AI approaches represent the most exciting frontier. Large datasets, such as those from the United Kingdom Biobank and Chronic Renal Insufficiency Cohort projects, comprising more than 100000 ECGs, have been used to train models specifically for artifact detection. These systems have achieved sensitivity and specificity rates exceeding 90% in distinguishing physiologic from nonphysiologic signals[14]. The major advantage of AI is its ability to rapidly flag poor-quality recordings, thereby saving clinicians valuable time. However, limitations remain: Current models cannot reliably detect electrode misplacement, and they cannot fully replace human interpretation when arrhythmia and artifacts overlap. Thus, AI can assist but not yet supplant the clinician. In the future, AI integration could revolutionize routine monitoring. Bedside monitors might soon provide real-time alerts for suspected artifacts, Holter recorders may automatically mark noisy segments for later inspection, and wearable devices might even prompt users to reposition electrodes when signal quality declines. These advancements have the potential to increase diagnostic accuracy and minimize misinterpretation, especially in busy or resource-constrained environments.

FUTURE DIRECTIONS

Although artifacts can never be completely eliminated, future developments aim to make their management easier and more reliable. Advances in smarter electrodes, such as self-adhesive patches that continuously monitor skin impedance and generate alerts when contact is poor, could significantly reduce electrode-related noise. Improvements in wearable technology, including sweat-resistant materials and stronger adhesives, may improve the reliability of long-term ambulatory recordings. At the system level, integrating machine learning into clinical monitors will likely become standard, providing real-time artifact detection and prompting corrective actions. In addition to technology, simulation-based training is crucial. Virtual ECG simulators that expose trainees to common artifact patterns alongside genuine arrhythmias can enhance diagnostic skills and reduce misinterpretation in clinical practice. Indeed, such simulators closely align with current educational projects, which aim to bridge the gap between engineering and clinical cardiology.

CONCLUSION

ECG artifacts are common challenges in clinical practice and are capable of mimicking arrhythmias and ischemia, leading to unnecessary interventions. They occur because the cardiac signal is fragile and easily disrupted during recording. Recognition relies on logic: Always check the patient, assess physiological plausibility, analyze lead distribution and baseline behavior, and confirm with ancillary signals. Prevention depends on careful electrode preparation, equipment maintenance, environmental control, and staff training. While digital filtering and AI offer new tools, clinical reasoning remains essential. The key message is simple: Never treat the ECG until you have checked the patient. Always remember, “The brain filters what machines cannot: Patient safety demands it”.

ACKNOWLEDGEMENTS

All authors contributed equally to this manuscript. We thank Dr. P Hema Krishna (neurologist) and Dr. Stalin for providing the deep-brain stimulator artifact and the square wave artifact electrocardiograms, respectively. We thank our Cath-lab staff for their unwavering technical support while collecting the best possible images. We thank Mr. Salehin Anjum for making the audio clip for the Core Tip.