What the vector loop shows,
and how to read it

The vector loop (vectorcardiogram, VCG) view turns a 12-lead ECG into a picture of where the heart's electrical activity points in space, and how that direction changes through a beat.

This manual explains what a vector loop is, how to read one, how EKGLab derives it, and what the tool will and will not tell you.

1What a vector loop is

At any instant during a heartbeat, the electrical activity of the whole heart can be summarised as a single vector: an arrow with a direction and a length. The direction is where the depolarization or repolarization wavefront is heading. The length is how much muscle is active at that moment.

That vector is not static. It swings through the beat as different regions activate. If you plot the tip of the arrow moment by moment, the tip traces out a closed path. That path is the vector loop.

A standard 12-lead ECG shows the same electrical activity as twelve separate traces of voltage against time. A vector loop shows it as one object in space, where direction and magnitude are visible directly and time is implied by the order the loop is drawn.

Each beat produces three loops, one per phase:

The QRS loop is normally a smooth, narrow oval, and it points in the same direction as the cardiac electrical axis.

2Reading the loop

The origin

The centre of the display is the electrically neutral point. Everything is measured from there.

Direction and magnitude

For any point on the loop, draw a line from the origin to that point. That line is the heart vector at that instant: its direction is where the current is heading, and its length is the magnitude in millivolts.

So a loop that reaches far to the lower left means the heart vector at that moment was large and directed leftward and inferiorly.

The timing dots

Dots are placed along the trace at a fixed interval of 10 ms. Because the interval is fixed, the spacing between dots tells you how fast the vector is moving:

• Dots spread far apart: the vector is changing rapidly.
• Dots crowded together: the vector is changing slowly.

Within the QRS loop, crowding reflects slowed conduction through the myocardium. This is the basis of a classic vectorcardiographic observation: in bundle branch block the terminal portion of the QRS loop shows closely spaced dots, because the last part of the ventricle to activate does so through slow muscle-to-muscle conduction rather than through the rapid conduction system.

The arrows

Arrows show the order in which the loop is inscribed, running P, then QRS, then T. They tell you where the loop starts and which way it travels.

A warning about rotation direction

Classical vectorcardiography texts describe the normal direction of rotation of the QRS loop within each plane, for example clockwise or counterclockwise. Those descriptions depend entirely on which side of the plane you are viewing from, because viewing a plane from the opposite side reverses the apparent rotation.

This tool lets you rotate the view freely. A loop that appears to run clockwise will appear to run counterclockwise once you have dragged past it. For that reason this manual does not give normal rotation directions: in a freely rotatable view they would be true only for some camera positions and false for the rest.

If you want to assess rotation, use the plane buttons rather than a hand-dragged view, so that you know exactly which direction you are looking from, and treat the result with the same care you would give any viewpoint-dependent sign.

Relationship between the QRS and T loops

In a normal heart, depolarization and repolarization run in broadly the same direction, so the QRS and T loops point broadly the same way. When they diverge widely, that discordance accompanies conduction abnormality, ventricular hypertrophy and ischaemia.

The vector loop view shows you this relationship visually. It does not compute a spatial QRS-T angle. See section 8.

3The three planes

A vector loop is a three-dimensional object. Classical vectorcardiography reads it as three flat projections, one per anatomical plane.

The axes follow the Frank convention:

The three planes are formed from pairs of those axes:

These are not three different recordings. They are three views of the same loop. In EKGLab the plane buttons rotate the camera to each of these orientations, so you can watch one view turn into another and see directly that they are projections of a single object.

4How EKGLab derives the loop

Measured versus derived

Classical vectorcardiography records X, Y and Z directly, using the Frank orthogonal lead system: seven electrodes placed on the chest, back, flanks, neck and left leg, wired through a resistor network so that three orthogonal leads come straight off the patient.

Frank recordings are rarely made in modern clinical practice, because they need electrodes that a standard 12-lead does not. Instead, X, Y and Z are derived mathematically from the 12-lead ECG that was recorded anyway.

EKGLab derives the loop. It does not measure it. Every loop you see in this tool is a mathematical reconstruction from the 12-lead.

The transformation

EKGLab uses the Kors regression transformation. Each orthogonal lead is a weighted sum of eight of the standard leads (I, II and V1 through V6; the remaining limb leads are linear combinations of I and II and so add nothing).

Why this transformation

Several transformations exist. Jaros, Martinek and Danys compared four of them against directly measured Frank leads and reported the median correlation between derived and measured signals:

Kors regression was the most accurate on all three axes, and significantly so for X and Y.

For contrast, the simplest alternative is the Kors quasi-orthogonal method, which just takes X = V6, Y = II and Z = −0.5 × V2. It is far easier to compute and measurably worse, particularly on Z.

An important qualification on those figures. They were established on the PTB database. EKGLab's library is PTB-XL, which contains a far higher proportion of pathological recordings, and the authors name validation on pathological records as work still to be done. Treat the correlations above as evidence that Kors regression is the right choice of transformation, not as an accuracy figure for any individual loop you are looking at.

5Using the tool

Opening it

Load a PTB-XL record, open the 12-lead panel, and press Vector loop. The view needs a 12-lead record; it cannot be built from a single-lead recording.

Moving the view

On opening, the loop turns through one revolution and then stops. That single rotation is there to make the shape readable as a three-dimensional object; after it, nothing moves unless you move it.

Arriving at a plane

You do not have to use the buttons. If you drag the loop to within a few degrees of a named plane, the view eases the rest of the way, that plane's button lights up, and the plane's own instrument appears. Drag away and it releases.

When the camera is between planes, no button is lit, because you are not looking at any named projection.

Gain

The gain control in the 12-lead header sets the scale for the loop, from 2.5 to 40 mm/mV, exactly as it does for the ECG grid.

All three loops share one gain, as they do on a real vectorcardiograph, where a single beam draws the whole beat and no part of it can be scaled separately. This is why the P loop looks small: it genuinely is small.

To inspect the P loop, raise the gain. The QRS will grow past the edge of the frame and be clipped. That is not a fault. It is what a vectorcardiographer does when they want to see the atrial loop, and what the instrument does in response.

The reference box and calibration

The 3D view is drawn inside a gridded box that is a fixed size on screen. It is a ruler, not part of the data. The readout in the corner tells you what one grid division is worth at the current gain, for example 1 division = 1.00 mV at 10 mm/mV.

Because the box never changes size, a low-voltage record renders as a genuinely small loop, which is itself diagnostic information. The loop is not scaled up to fill the frame.

6Colours and labels

Phases

Anatomical axes

In the 3D view, three arrows radiate from the origin and are labelled LEFT, INFERIOR and POSTERIOR. These are the positive directions of X, Y and Z. Because the arrows rotate with the loop, they stay correct at every angle, and the opposite direction of each arrow is right, superior and anterior respectively.

The frontal wedge

When the view is at the frontal plane, a shaded wedge appears covering −30° to +90°. That is the normal range for the QRS electrical axis. A QRS loop whose long axis falls inside the wedge has a normal axis; outside it, to the left of −30° is left axis deviation and beyond +90° is right axis deviation.

The wedge appears only on the frontal plane, because the QRS axis is defined in that plane and the range means nothing in the others.

7What the tool tells you about itself

The vector loop reports its own working, and it is worth reading.

Beat averaging

The loop is not a single beat. It is the average of the beats in the recording that share the dominant QRS morphology. The caption says how many contributed, for example Representative of 8 beats.

Averaging suppresses noise, at the cost of hiding beat-to-beat variation.

Excluded beats

Beats whose QRS morphology differs markedly from the dominant cluster, such as ectopics, are excluded from the average and drawn separately in amber. The caption reports them, for example 2 outliers excluded. If that count is high, the recording is not dominated by a single morphology and the representative loop means less.

Estimated phase boundaries

The P, QRS and T colouring depends on knowing where each phase begins and ends. Those boundaries come from measured values when they are available. When they are not, the tool falls back to typical intervals and appends phases estimated to the caption. The loop shape is unaffected; only the colour boundaries are approximate. A caption reading approximate means the measured boundaries were usable but had to be adjusted.

Absent phases

If a phase is too small to be distinguished from baseline noise, it is not drawn as a coloured loop, and the readout says P absent or T absent.

This matters most in atrial fibrillation, where there is no organised atrial depolarization and therefore no P loop. Rather than magnify whatever noise occupies the P window into a convincing-looking loop, the tool declines to draw one and says so. Absence of a P loop is information, not a rendering failure.

8Limitations

The loop is derived, not measured. It is a mathematical reconstruction from the 12-lead, not a Frank recording. Section 4 covers the accuracy of that reconstruction and the population it was established on.

No spatial QRS-T angle. The vector loop does not compute one. The QRS-T angle shown in the cardiac axis panel is the frontal-plane angle, the difference between the frontal QRS and T axes. The spatial QRS-T angle, the true three-dimensional angle between the mean QRS and T vectors, is a different measurement with different thresholds, and it is the one carrying the prognostic literature. Do not read one for the other.

No loop metrics. The tool does not compute loop area, planarity, maximum vector magnitude or any other quantitative descriptor. It is a viewer.

Rotation direction is viewpoint dependent. See section 2.

One averaged beat. Beat-to-beat variability is not shown, other than through the excluded-beat count and the superimposed traces.

9Scope and clinical disclaimer

The vector loop view is a teaching tool. It is intended to help students and clinicians develop an intuition for cardiac electrical activity in space, and to support instruction alongside the 12-lead ECG.

It is not a diagnostic device, and nothing it displays should be used to make or exclude a clinical diagnosis. It computes no diagnostic verdict. The colours, the wedge and the calibration exist to help you read the picture, not to tell you what it means.

Recordings in the EKGLab library come from published research databases and are provided for education.

10Reference

Jaros, R.; Martinek, R.; Danys, L. Comparison of Different Electrocardiography with Vectorcardiography Transformations. Sensors 2019, 19, 3072. doi:10.3390/s19143072

The transformation coefficients in section 4 are that paper's Table 2. The correlation figures are its Table 6. The axis conventions in section 3 follow its description of the Frank orthogonal lead system.