Unmasking concealed arrhythmia syndromes

Diagnosing inherited arrhythmia syndromes is difficult in patients and relatives even after they have experienced an arrhythmia, syncope or survived cardiac arrest. The difficulty is greatest when the resting ECG is also normal (concealed phenotype).

Transient electrical manifestations during follow-up may be the only clue, but ambulatory ECG monitoring is not a currently viable solution because of limited electrode coverage, limited duration and impracticality of manual reviews of multi-day ECG data.

My ambition is to combine the best of human expertise and extended-duration multi-electrode ambulatory ECG monitoring, using machine learning and advanced ECG signal processing techniques, to unmask these concealed arrhythmia syndromes.

To achieve this, I will: - Develop two complementary machine learning algorithms to detect Brugada ECGs: one trained to match the expert consensus, another trained to measure the beta angle of the QRS complex. - Test these algorithms prospectively in participants wearing a novel multi-electrode ECG garment over a 3-month period: 50 patients with manifest Brugada and 50 healthy volunteers. - Apply the algorithms in 50 patients with concealed Brugada syndrome to look for episodes of Brugada pattern unmasking. - Apply my developed algorithms for Brugada, and existing algorithms for long QT syndrome, in 50 idiopathic VF patients.