Optimal length of R–R interval segment window for Lorenz plot detection of paroxysmal atrial fibrillation by machine learning

Abstract

Abstract Background Heartbeat interval Lorenz plot (LP) imaging is a promising method for detecting atrial fibrillation (AF) in long-term monitoring, but the optimal segment window length for the LP images is unknown. We examined the performance of AF detection by LP images with different segment window lengths by machine learning with convolutional neural network (CNN). LP images with a 32 × 32-pixel resolution of non-overlapping segments with lengths between 10 and 500 beats were created from R–R intervals of 24-h ECG in 52 patients with chronic AF and 58 non-AF controls as training data and in 53 patients with paroxysmal AF and 52 non-AF controls as test data. For each segment window length, discriminant models were made by fivefold cross-validation subsets of the training data and its classification performance was examined with the test data. Results In machine learning with the training data, the averages of cross-validation scores were 0.995 and 0.999 for 10 and 20-beat LP images, respectively, and > 0.999 for 50 to 500-beat images. The classification of test data showed good performance for all segment window lengths with an accuracy from 0.970 to 0.988. Positive likelihood ratio for detecting AF segments, however, showed a convex parabolic curve linear relationship to log segment window length and peaked at 85 beats, while negative likelihood ratio showed monotonous increase with increasing segment window length. Conclusions This study suggests that the optimal segment window length that maximizes the positive likelihood ratio for detecting paroxysmal AF with 32 × 32-pixel LP image is 85 beats.

Keywords

• Artificial intelligence
• Atrial fibrillation
• Convolutional neural network
• Holter electrocardiogram
• Lorenz plot
• Machine learning