Biological measurement

Charge, separated and then released

A resting muscle cell keeps ions unequally distributed across its membrane, pumping some out and holding others in. The separated charge maintains a potential difference across the membrane, the resting potential, with the inside about 90 mV negative to the outside. Nothing is flowing; the cell is holding a charge apart, like a very small charged capacitor.

Trigger the cell and that arrangement collapses in a few milliseconds. Ions cross the membrane, the p.d. briefly reverses, and then the pumps restore it. The pulse is the action potential: the collapse is depolarisation and the recovery is repolarisation. One depolarising cell triggers its neighbours, so what actually happens in a heartbeat is a wave of depolarisation sweeping across the muscle.

That sweep is what an electrocardiograph can see. Part of the heart is depolarised and part is not, so at every instant there is a separation of charge with a size and a direction, and the whole heart behaves as an electric dipole that grows, turns and shrinks through the beat. Body fluids conduct, so that dipole sets up small potential differences right across the torso. Put two electrodes on the skin and the p.d. between them is of the order of a millivolt, and it changes shape in step with the beat.

So the opening claim is wrong twice over. The trace plots potential difference against time, not motion, and the electrical event is the cause of the contraction rather than a record of it. Muscle contracts shortly after it depolarises.

Getting a millivolt out

The signal is small, the source is poor, and the room is full of interference, so an ECG machine is built around those three problems.

It starts with electrodes stuck to the skin, at least two to define a p.d. and in practice several so that the dipole can be viewed from different directions. Dry skin is a poor contact with a resistance of tens of kilohms, so a conducting gel goes between electrode and skin to bring that down and to keep the reading steady as the patient moves.

The amplifier then has three requirements worth learning as a list. It needs high gain, because a millivolt has to become a signal of a volt or so before anything can display it. It needs a very high input impedance, because the electrode contact is itself a large resistance: draw any appreciable current through it and most of the signal is lost across the contact instead of reaching the amplifier.

The third requirement is low noise. The amplifier works on the difference between two electrodes and rejects whatever they both pick up, which is chiefly 50 Hz mains hum arriving equally at every electrode on the patient. Filtering removes what survives, and the result is displayed on a screen or printed on moving paper.

Reading a normal trace

One beat produces three features in a fixed order, and each is one stage of the depolarisation sweep.

The P wave is the depolarisation of the atria, the two upper chambers, which then contract and push blood into the ventricles. It is small, a few tenths of a millivolt, because the atria carry little muscle.

The flat stretch after it is the delay while the signal is held up on its way to the ventricles, giving them time to fill. Nothing much is depolarising, so the trace sits near the baseline for something like 0.15 s.

The QRS complex is the depolarisation of the ventricles, the two pumping chambers, which then contract and drive blood out to the lungs and the body. It is by far the largest deflection, about a millivolt, because the ventricles carry much more muscle than the atria, and it is narrow because that mass depolarises quickly and almost together. The atria repolarise during it, and their much smaller signal is buried underneath.

The T wave is the repolarisation of the ventricles as they recover and relax. It is lower and broader than the QRS because repolarisation is a slower process and less well synchronised across the muscle, so the same recovery is spread over more time.

The rate comes off the trace by measuring between two features that are easy to find, and the tall R spike is the obvious choice. An R to R interval T seconds long means 60/T beats each minute. Calling a trace normal is then a judgement about four things at once: the three features present and in order, their relative sizes as described above, the intervals within their usual ranges, and the beats evenly spaced. A trace is read by comparing it against that, feature by feature.