Introduction: This study sought to develop and validate diagnostic models to identify individuals with atrial fibrillation (AF) using amplified p-wave analysis or neural-network-trained p-wave-analysis during sinus rhythm (SR).

Methods and results: 1492 patients (491 healthy controls, 499 with paroxysmal AF and 502 with persistent AF) underwent digital 12-lead-ECG recording during sinus rhythm. The patient cohort was divided into training and validation set in a 3:2 ratio. P-wave indices including duration of standard p-wave (standard PWD; 10mm/mV, 25mm/sec, see ECG in the left panel), amplified digital p-wave (APWD, 60-100mm/mV, 175mm/sec; see ECG in the middle panel) and advanced inter-atrial block (aIAB) along with the parameters body surface area (BSA) and LA diameter (LAD) were used to develop diagnostic models using logistic regression. Each model was developed from the training set and further tested in both training and validation sets for its diagnostic performance in identifying individuals with AF. A neural-network was trained to identify the p-wave and its features (duration, morphology) to differentiate between individuals with known versus those without AF.

Compared to standard PWD (Reference model, AUC=0.637 and 0.632, for training and validation set, respectively), APWD (Basic model) importantly improved the accuracy to identify individuals with AF (AUC=0.86 and 0.866). The ECG-based model combining APWD, aIAB and body surface area (BSA) further improved the diagnostic performance for AF (AUC=0.892 and 0.885). The integrated model, which further combined left atrial diameter (LAD) with parameters of the ECG-based model, achieved the highest diagnostic performance (AUC=0.916 and 0.902). Notably, the neural network-based p-wave-analysis identified AF patients with high accuracy during sinus rhythm (AUC: 0.85; sensitivity: 77%; specificity: 78%).

Conclusions: Analysis of amplified p-wave or neural-network-based automatic p-wave analysis during sinus rhythm allow identification of individuals with atrial fibrillation. The current diagnostic models provide a new basis for identifying individuals at high risk for AF, who would most benefit from AF screening.