Algorithm for Recognition of Left Atrial Appendage Boundaries in Echocardiographic Images

Abstract

Introduction: The left atrial appendage )LAA( occlusion using a purpose-built device is a growing procedure. This study aimed to develop a computer-aided diagnostic system for the recognition of the LAA in echocardiographic images. Material and Methods: The three-dimensional (3D) echocardiographic images of the LAA of 26 patients successfully treated with an LAA occluder were used in this study. A total of 208 3D derived two-dimensional images in the axial plane were derived from each 3D dataset. Then, 562 images in which the LAA boundaries were highly recognizable were selected for this study. The proposed convolutional neural network (CNN) in this study was based on open-source object identification and classification platform compiled under the You Only Look Once algorithm. Finally, 419 and 143 images were used for training and testing the algorithm, respectively. Results: Algorithm performance on the identification of the LAA region on a set of 143 images was compared to that reported for the traced regions on the same images by an expert in echocardiography using an intersection over the union (IOU) algorithm. The algorithm was able to correctly identify the LAA region in all 143 examined images with an average IOU of 90.7%. Conclusion: The proposed image-based CNN algorithm in this study showed high accuracy in the recognition of the LAA boundaries in the echocardiographic images. The method can be used in the development of algorithms for the automated analysis of the area of the LAA used for device sizing and procedural planning in the LAA occlusion procedures.

Keywords

• artificial intelligence
• atrial fibrillation
• computer vision
• echocardiographic
• left atrial appendage