Automated three-dimensional left atrial analysis on computed tomography angiography: reproducibility and workflow efficiency of eight clinically relevant metrics using a deep learning pipeline

TL;DR

This study introduces a deep learning pipeline that automates the analysis of left atrial anatomy in CT scans, showing high accuracy and efficiency compared to manual methods.

Contribution

A novel deep learning pipeline for automated left atrial analysis on CTA is validated for clinical metrics with expert-level accuracy and efficiency.

Findings

Automated measurements showed excellent agreement with experts for LA volume and diameters (ICC = 0.999–0.985).

Analysis time was reduced by 92% compared to manual methods (0.5 min vs. 15.3 min per case).

Usability ratings were high (≥4/5) and 63–76% of cases were classified as fully usable without edits.

Abstract

Accurate and reproducible quantification of LA anatomy from CTA is essential for ablation, LAAC, and structural interventions, yet manual measurements are time-consuming and prone to inter-observer variability.

To validate a deep learning–based CTA pipeline for automated quantification of eight clinically relevant LA metrics, and to assess its agreement, repeatability, and efficiency compared with expert measurements.

In this retrospective study, 407 patients were included and divided into training (n = 270), validation (n = 87), and clinical evaluation (n = 50) cohorts. A MedNeXt-based model performed multi-structure segmentation, and a geometry-driven framework computed eight metrics: LA volume, LAA volume, AP/ML/SI diameters, left/right PV ostial size, left/right PV inter-ostial angles, and LAA ostial size. Automated outputs were compared with expert annotations using intraclass correlation coefficients (ICCs) and Bland–Altman analysis (primary endpoints: LA volume and diameters). Workflow efficiency and usability (Likert scale, 1–5) were also assessed by electrophysiology/structural experts.

Automated measurements demonstrated excellent agreement with experts for primary endpoints (LA volume ICC = 0.999; AP/ML/SI diameter ICCs = 0.972–0.985), with minimal bias and narrow limits of agreement. Agreement for other metrics was good to excellent (typical ICCs ≥0.84). Analysis time was reduced from 15.3 ± 1.4 min to 0.5 ± 0.1 min per case (≈92% reduction; p < 0.001). Usability ratings were ≥4/5 in most cases, with 63%–76% classified as Grade A (fully usable without manual edits). Performance remained consistent across voxel-size strata.

The proposed pipeline enables rapid, reproducible, and expert-level quantification of eight LA metrics on CTA, demonstrating technical feasibility for clinical integration pending prospective validation of procedural impact.

Figures