A cardiograph is a device that records the electrical activity of the heart, producing a tracing called an electrocardiogram (commonly abbreviated ECG or EKG). Every heartbeat is triggered by a tiny electrical impulse that spreads through heart muscle in a predictable pattern, and a cardiograph detects those impulses through sensors placed on the skin. The technology dates back more than a century, but the core principle remains the same: pick up faint voltages at the body’s surface and translate them into a waveform a clinician can read.
How a Cardiograph Picks Up Your Heartbeat
Your heart’s electrical signals begin in a small cluster of cells in the upper right chamber and ripple outward through the muscle. Those signals are strong enough to reach the skin, though by the time they arrive they have weakened to roughly one millivolt or less. A cardiograph captures them using electrodes, typically sticky patches with a conductive gel, placed at specific points on the chest and limbs. The electrodes act as transducers, converting the ionic current flowing through body tissue into electronic current that the machine’s circuitry can process.1PubMed Central. Human Body-Electrode Interfaces for Wide-Frequency Sensing and Communication: A Review
A standard clinical ECG uses ten electrodes to generate twelve different “leads,” which are really just twelve different vantage points on the heart’s electrical field. Six electrodes go on the chest, and four go on the wrists and ankles. Each lead compares the voltage between two points (or between one point and a calculated average), giving clinicians a view of different walls of the heart. If one region is damaged or conducting electricity abnormally, that abnormality tends to show up in the leads facing that region.
Because the signals are so small, the circuitry inside the machine has to amplify them enormously while filtering out noise from skeletal muscles, power lines, and other electrical interference. Wearable ECG devices face an even tougher version of this challenge, since the electrodes may not sit as snugly and the wearer keeps moving. Researchers have developed circuits specifically aimed at canceling the common-mode noise that creeps in when contact between skin and electrode is uneven, reducing interference to levels low enough for a usable reading.2PubMed Central. Common-mode noise cancellation circuit for wearable ECG
What the Waves on an ECG Actually Mean
The familiar up-and-down tracing that a cardiograph produces is not random squiggles. Each heartbeat creates a repeating pattern of waves labeled P, Q, R, S, and T. These letters have stuck around since the late 1800s and early 1900s, when Willem Einthoven refined the first practical electrocardiograph. Einthoven built on earlier physiologists’ work studying the heart’s electrical behavior and eventually won the Nobel Prize in 1924 for turning the concept into a clinical instrument.3PubMed Central. Einthoven’s string galvanometer: the first electrocardiograph
Each wave corresponds to a specific phase of the heartbeat. The P wave represents the electrical spread through the upper chambers (atria), causing them to squeeze. The QRS complex, a sharp spike, marks the moment the lower chambers (ventricles) are activated and contract forcefully to pump blood. The T wave reflects the ventricles resetting their electrical charge, getting ready for the next beat. The flat segments between waves matter too. The interval between the P wave and the QRS complex tells you how quickly the signal is traveling from the upper chambers to the lower ones, while the segment between the QRS and the T wave reflects conditions inside the ventricular muscle itself.
What Doctors Look for on an ECG
Clinicians read an ECG the way a mechanic listens to an engine. They check heart rate, rhythm regularity, wave shapes, and the timing between waves. A normal ECG has a predictable rhythm, consistent wave shapes, and intervals that fall within known ranges. Deviations from those norms point toward specific problems.
Heart rhythm disorders, or arrhythmias, are among the most common reasons for ordering an ECG. Sinus node dysfunction, for example, can produce a range of ECG patterns including unusually slow heart rates, long pauses between beats, and sudden switches between slow and fast rhythms.4Revista Española de CardiologÃa (English Edition). Bradyarrhythmias and Conduction Blocks The list of diagnosable rhythms is long. A deep-learning study trained its algorithm on 21 distinct rhythm types detectable from ECGs, ranging from normal sinus rhythm to atrial fibrillation, various degrees of heart block, and pre-excitation syndromes.5The Lancet Digital Health. Automated multilabel diagnosis of heart rhythm or conduction abnormalities from electrocardiograms using a deep learning approach: a multicentre study
Heart attacks leave distinctive footprints on the tracing as well. During a heart attack, the injured muscle generates abnormal electrical currents that show up as characteristic shifts in the ST segment, the flat line between the QRS complex and the T wave. Research using heart simulations has shown that under moderately ischemic conditions, differences in ion channel behavior across the heart wall create a voltage gradient that pushes the ST segment upward on the ECG.6PubMed Central. Ionic mechanisms of ST segment elevation in electrocardiogram during acute myocardial infarction At the molecular level, the opening of certain potassium channels in oxygen-starved cells is a key driver of that ST elevation.7PubMed. Molecular basis of electrocardiographic ST-segment elevation The surface ECG changes in the leads facing the injured area reflect injury currents flowing across the boundary between healthy and ischemic tissue.8PubMed Central. Acute myocardial ischemia: cellular mechanisms underlying ST segment elevation Recognizing these patterns quickly can be the difference between opening a blocked artery in time and suffering permanent heart damage.
When a Snapshot Is Not Enough
A standard 12-lead ECG records about ten seconds of heart activity. That works well for diagnosing ongoing problems, but many arrhythmias are intermittent. You might feel your heart race or skip once a week, and the odds of it happening during your brief clinic visit are slim. That is where ambulatory monitors come in.
A Holter monitor is a portable device you wear for 24 to 48 hours while going about your daily life. It records every heartbeat during that window so a technician can scroll through it later. For problems that happen even less often, longer-term devices exist: external loop recorders can be worn for weeks, and implantable loop recorders sit under the skin and record for years. A health technology assessment found that both long-term external monitors and external loop recorders were more effective at detecting symptomatic arrhythmias than a standard 24-hour Holter, with no substantial difference between the two longer-term approaches.9PubMed Central. Long-Term Continuous Ambulatory ECG Monitors and External Cardiac Loop Recorders for Cardiac Arrhythmia: A Health Technology Assessment
Newer wearable monitors have pushed ambulatory monitoring further. One device designed to record a 72-hour triple-lead ECG with real-time cloud transmission showed no significant differences from traditional Holter monitoring in total heart rate, average heart rate, and counts of premature beats. Its sensitivity for detecting premature atrial and ventricular complexes, as well as atrial fibrillation, ranged from 93 to 98 percent, with specificity of 98 to 99 percent.10PubMed Central. A wearable real-time telemonitoring electrocardiogram device compared with traditional Holter monitoring The practical upside is that a doctor can review your rhythm remotely while you stay at home, rather than requiring an in-person visit to retrieve data.
Things That Can Fool the Machine
An ECG is only as clean as the signal reaching the electrodes. Artifacts, meaning distortions that have nothing to do with your heart’s electrical activity, are a persistent headache in clinical practice. They can warp the baseline, distort wave shapes, or mimic arrhythmias that are not actually happening.
Motion is the most common culprit. Skeletal muscles generate their own electrical signals, and when they fire near an electrode the ECG gets “bombarded” by apparently random activity. Tremors from conditions like Parkinson’s disease, anxiety, hypothermia-related shivering, or stimulant medications can all produce motion artifacts.11PubMed Central. Main artifacts in electrocardiography Even simple limb movement during the test can create sudden baseline shifts that look like premature heartbeats or other arrhythmias. Loose or dried-out electrode patches, electrical interference from nearby equipment, and incorrect lead placement add to the list. Experienced clinicians learn to recognize artifacts, but they occasionally slip through, especially in automated readings. This is one reason a computer-generated ECG interpretation almost always includes the disclaimer “confirmed by physician.”
How Computers Read ECGs
Automated ECG interpretation has been around since the 1960s, but it has improved dramatically. The first step in any automated reading is detecting the QRS complex, that sharp spike in each heartbeat, since everything else is measured relative to it. Classical detection algorithms use the rate of change in the signal (the first derivative) to find QRS peaks. On standard test databases, the best-performing algorithms achieve sensitivities above 99.5 percent, though errors still cluster around beats with unusual shapes, such as wide arrhythmic beats or low-amplitude signals.12PubMed Central. Analysis of first-derivative based QRS detection algorithms
More recently, deep-learning models have been trained not just to find individual heartbeats but to classify entire rhythm strips into dozens of categories simultaneously.5The Lancet Digital Health. Automated multilabel diagnosis of heart rhythm or conduction abnormalities from electrocardiograms using a deep learning approach: a multicentre study And the ambitions keep growing. Researchers have developed AI algorithms that use routine ECG data, combined with echocardiographic outcomes, to screen for structural heart conditions like hypertrophic cardiomyopathy, conditions you would normally need an ultrasound to find. Early results have been promising enough to suggest that the humble ECG could evolve from a rhythm-and-conduction tool into a broader screening instrument for heart structure and function.13PubMed Central. Artificial Intelligence in Electrocardiography: From Automated Arrhythmia Detection to Predicting Hidden Cardiovascular Disease
Smartwatches and Consumer ECGs
If you own a recent Apple Watch or certain other smartwatches, you already carry a simplified cardiograph on your wrist. These devices use a single-lead ECG recorded between an electrode on the watch’s back (touching your wrist) and a second electrode you touch with a finger on the opposite hand. The recording takes about thirty seconds and is paired with an app that automatically classifies the rhythm, with a particular focus on detecting atrial fibrillation.14PubMed. How useful is the smartwatch ECG?
Some smartwatches also use photoplethysmography, which shines a green light into the skin and measures blood-volume changes to estimate heart rate continuously. When researchers validated one such device’s combined optical and ECG algorithm against a 28-day continuous ECG patch, the smartwatch performed well enough in a free-living setting to be considered a useful screening tool for atrial fibrillation.15PubMed. Validation of an algorithm for continuous monitoring of atrial fibrillation using a consumer smartwatch
That said, a single-lead consumer recording captures far less information than a 12-lead clinical ECG. It is good for rhythm monitoring but cannot reliably detect heart attacks, conduction blocks involving specific parts of the heart, or many structural abnormalities. Think of it as a screening tool that tells you when to see a doctor, not a replacement for one.
Beyond Electrical Signals: Impedance Cardiography
Not every cardiograph measures electrical activity. Impedance cardiography (ICG) measures how much the chest’s resistance to a tiny electrical current changes with each heartbeat. When the heart pumps blood into the aorta, the thorax’s electrical impedance shifts briefly, and those shifts can be used to calculate stroke volume, the amount of blood ejected with each beat.16PubMed Central. Design and Implementation of a Portable Impedance Cardiography System for Noninvasive Stroke Volume Monitoring The technique is noninvasive, relatively inexpensive, and suited for continuous monitoring, which makes it appealing in situations where invasive catheterization is impractical.
Validation studies have compared impedance-derived stroke volume against established methods. In one study of patients with high blood pressure, impedance cardiography correlated well with dye-dilution measurements, with correlation coefficients between 0.81 and 0.92 regardless of which calculation formula was used.17PubMed. Stroke volume measurement by impedance cardiography using a formula based on the delta z waveform During exercise, impedance estimates tracked reasonably well against COâ‚‚ rebreathing-based measurements from rest up to about 70 percent of peak effort, though the correlation was moderate rather than tight.18PubMed. Stroke volume measurement during supine and upright cycle exercise by impedance cardiography ICG is not a replacement for high-precision methods in critical-care settings, but it fills a useful niche for trend monitoring and hemodynamic screening without needles or catheters.
Vectorcardiography and Three-Dimensional Views
A standard ECG shows voltage changes over time, one lead at a time. Vectorcardiography (VCG) takes a different approach: it plots the heart’s electrical signal in three-dimensional space, showing the direction and magnitude of the electrical force at every moment during the heartbeat. VCG signals monitor both spatial and temporal cardiac electrical activity along three orthogonal planes of the body.19PubMed Central. Spatiotemporal representation of cardiac vectorcardiogram (VCG) signals The result is a set of loops rather than a series of waves. Clinicians can see, for instance, that the main electrical force during ventricular activation points down and to the left (toward the body’s largest heart chamber), and deviations from the normal loop shape or orientation can reveal conduction problems, enlarged chambers, or past heart damage.
VCG never replaced the standard 12-lead ECG in routine practice, partly because 12-lead interpretation is deeply ingrained in clinical training and guidelines. But VCG data can be mathematically derived from a standard 12-lead recording, and some modern analysis software generates VCG loops alongside traditional tracings. Researchers continue to find situations where VCG adds diagnostic value, especially when subtle changes in the three-dimensional orientation of electrical forces are hard to spot on flat waveform printouts.
Fetal Heart Monitoring
The same principle that works for adults can be adapted to monitor a baby’s heartbeat before birth. Non-invasive fetal electrocardiography places electrodes on the mother’s abdomen and extracts the fetal signal from a noisy mix of maternal ECG, muscle activity, and movement artifacts. Separating the baby’s tiny signal from the mother’s much larger one requires specialized signal processing, but advances in filtering have made the technique reliable enough to provide fetal heart rate data comparable to what an invasive scalp electrode delivers during labor.20PubMed Central. Non-invasive Fetal Electrocardiography for Intrapartum Cardiotocography The advantage over Doppler ultrasound, the standard non-invasive method, is that an electrode-based approach can provide beat-to-beat timing information and potentially detect fetal arrhythmias that ultrasound would miss.
ECGs in Other Species
Cardiography is not limited to humans. Veterinarians routinely use ECGs on dogs, cats, horses, and other animals, though lead placement and normal wave patterns vary considerably by species. Comparative research has revealed fundamental differences in how hearts work across the animal kingdom. In mammals and birds, heart chamber activation and repolarization are much faster than in cold-blooded vertebrates of similar size. Reptiles measured at the same body temperature as mammals had roughly four-fold lower heart rates, two-fold slower conduction through the atria and ventricles, and four-fold longer intervals for atrioventricular delay and total ventricular repolarization.21PubMed. The electrocardiogram of vertebrates: Evolutionary changes from ectothermy to endothermy The researchers concluded that temperature alone cannot explain the speed difference; warm-blooded animals have more compact heart muscle architecture that supports faster electrical conduction. These findings underscore how the same cardiographic tool can illuminate heart physiology across wildly different bodies, from a hummingbird to a python.