Towards Patient-Specific Surgical Planning for Bicuspid Aortic Valve Repair: Fully Automated Segmentation of the Aortic Valve in 4D CT

2025-02-13

AI Summary

Bicuspid aortic valve (BAV) repair lacks personalized preoperative planning tools. Method: This study introduces the first fully automated, multi-label 4D CT segmentation framework specifically designed for BAV, built upon the nnU-Net architecture with novel optimizations for temporal data preprocessing and postprocessing to accurately delineate leaflets, commissures, and root structures. Results: In a clinical cohort, the framework achieved mean Dice scores >0.7 and symmetric mean surface distances <0.7 mm for leaflets and root walls. Key surgical metrics—including cusp height, commissural angle, and annular diameter—showed excellent agreement with manual annotations (ICC >0.92). This work presents the first systematic clinical usability validation of such a tool, enabling patient-specific geometric modeling and surgical risk stratification, thereby significantly enhancing objectivity and reproducibility in preoperative assessment for BAV repair.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Calibration & Uncertainty Quantification
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Security and Privacy: Large-scale security measurements

Abstract

The bicuspid aortic valve (BAV) is the most prevalent congenital heart defect and may require surgery for complications such as stenosis, regurgitation, and aortopathy. BAV repair surgery is effective but challenging due to the heterogeneity of BAV morphology. Multiple imaging modalities can be employed to assist the quantitative assessment of BAVs for surgical planning. Contrast-enhanced 4D computed tomography (CT) produces volumetric temporal sequences with excellent contrast and spatial resolution. Segmentation of the aortic cusps and root in these images is an essential step in creating patient specific models for visualization and quantification. While deep learning-based methods are capable of fully automated segmentation, no BAV-specific model exists. Among valve segmentation studies, there has been limited quantitative assessment of the clinical usability of the segmentation results. In this work, we developed a fully auto- mated multi-label BAV segmentation pipeline based on nnU-Net. The predicted segmentations were used to carry out surgically relevant morphological measurements including geometric cusp height, commissural angle and annulus diameter, and the results were compared against manual segmentation. Automated segmentation achieved average Dice scores of over 0.7 and symmetric mean distance below 0.7 mm for all three aortic cusps and the root wall. Clinically relevant benchmarks showed good consistency between manual and predicted segmentations. Overall, fully automated BAV segmentation of 3D frames in 4D CT can produce clinically usable measurements for surgical risk stratification, but the temporal consistency of segmentations needs to be improved.

Problem

• Automated segmentation of BAV in 4D CT.
• Quantitative assessment for surgical planning.
• Improving temporal consistency in segmentation.

Innovation

• Fully automated BAV segmentation pipeline
• Based on nnU-Net architecture
• Produces clinically usable measurements