Computer was taught to diagnose arrhythmia more accurately than a doctor

Atrial fibrillation (heartbeat rhythm failure) may be unnoticed on ECG, especially when the rhythm of the heartbeat returns to its normal condition for a while. According to BBC, artificial intelligence created by Mayo Clinic may help to avoid a fatal diagnostic error. Scientists have revealed arrhythmia signs hidden from the doctor’s eye.
The computer is looking for seemingly insignificant traces that speak of past episodes of heart rhythm failures. Such traces include the formation of scar tissue on the heart. The system analyzed heart rhythm studies of almost 181,000 people from 1993-2017. All people initially had normal tests. But artificial intelligence revealed arrhythmia in 83% of studies stated as “normal”.
There is no doubt in the power of artificial intelligence. For example, a computer recently developed a high-performance flu vaccine. The vaccine was created by a program known as Ligand Search Algorithm. Ligands are compounds (sometimes a small molecule) that form complexes with a biological molecule (DNA, protein, cell receptor) and, as a result, produce certain effects.

Source: BBC.