Change of heart diagnostics, numbers to make sense of ECG reports

We may not realise it but our hearts are like electric generators, inducing a small bit of electric current every time they beat. The electricity generated is a very small amount, of the order of a few millivolts, or a thousandth of a volt, enough to activate the heart functions.

The electric current generated by the heart can be used to assess the condition of the organ. In fact, doctors now use this as a test of the heart all the time, using an instrument called the electrocardiogram, or ECG for short. The ECG measures the electrical activity of the heart and produces a waveform on a graph paper. Doctors look at these waveforms to make an assessment of the health of the heart, whether it was beating normally or was under some kind of stress.

Unlike many other diagnostic tests, like those for blood samples or urine for example, ECG reports do not contain numerical values that can be compared to standard healthy values. In fact, what appears on the graph paper is not a single waveform. It is a combination of many waves of differing frequencies and amplitudes. That is because the heart is a very complex system.

Several things happen inside its different chambers every time it beats. The result is a complex non-linear ECG output, which an untrained person can make little sense of by just looking at it. Doctors use their expertise and experience to draw conclusions from these complex waveforms. The heights, or amplitudes, of these waveforms and the intervals at which they occur in the ECG output are a couple of important things that doctors look out for. There are several others.

What is more, there is no standard ECG waveform for a healthy heart. Different people produce very different ECG outputs. Age, gender, habits, stress or anxiety levels, and many other parameters determine the pattern that is produced. As such, doctors resort to a lot of correlational analysis with the patient’s personal and medical history to understand the ECG reports.

Though ECG reports have been quite a reliable tool in making an assessment of the health of the heart, scientists have been trying to extract more from these waveforms by looking at the complicated dynamics of the heart that produces these patterns on the graph paper. In particular, there is an attempt to explore whether it can be made more quantitative, and produce numerical values just like some other diagnostic tests do. The key to achieving this lies in identifying those set of numbers — the fewer the numbers in the set the better — out of the large number of parameters that the ECG instrument captures that can uniquely describe the condition of the heart.

This is what the research group led by G Ambika with members, Snehal M Shekatkar and Yamini Kotriwar, at the Indian Institute of Science Education and Research, Pune, and their collaborator, K P Harikrishnan at The Cochin College in Kochi, have managed to do under a Science and Engineering Research Board (SERB) research project funded by the Department of Science and Technology (DST).

Maintaining that a mere visual examination of the ECG reports does not reveal complete information on the complicated dynamics underlying the heart system, these scientists looked at data from hundreds of ECG reports, from healthy as well as diseased hearts.

Using different scientific and analytical tools, they were able to zero in on a set of four quantifiers whose values could uniquely reveal whether a heart was functioning normally or afflicted with some disease. Ambika says in fact just two of them were enough for this purpose. The range of values these quantifiers show and their variability for healthy hearts are very different from those of diseased hearts. It is therefore possible to distinguish between a diseased and normal heart just by knowing the values of these quantifiers, with a very high degree of accuracy.

As of now, with their current level of research, the scientists cannot zero in on the kind of disease that might be afflicting the heart by just looking at the values of these quantifiers. But Ambika says it is possible to do so. For that, they need to look at data from many more ECG reports for each specific disease.

The research, published in Scientific Reports last November, promises to significantly improve the capabilities of diagnosing heart diseases, and also allows for continuous monitoring of the heart for these specific quantifiers without the need to go in for full ECG reports.