Validation of Replicable Pipeline 3D Surface Reconstruction for Patient-Specific Abdominal Aortic Lumen Diagnostics

Accurate prognosis is challenging in high-risk vascular conditions such as abdominal aortic aneurysm, limited diagnostic standards, decisional criteria, and data semantics often hinder clinical reliability and impede diagnostics’ digital transition. This study aims to evaluate the performance, robustness, and usability of an automatic, replicable pipeline for aortic lumen surface reconstruction for pathologic vessels. The goal is to provide a solid tool for geometry reconstruction to a more complex enhanced diagnostics framework. Methods: A U-Net convolutional neural network is trained using pre-operative CTA scans, which 101 for model’s training and 14 for model’s testing, covering a wide anatomical and aorto-iliac pathologies spectrum. Validation included segmentation metrics, robustness, reliability, and usability assessments. Performance are investigated by means of test set prediction’s metrics, for several instances model’s input. Clinical reliability evaluates manual measurements performed by a vascular surgeon on the obtained 3D aortic lumen surfaces. Results: Test set is selected to cover a wide portion of aorto-iliac pathologies. The algorithm demonstrated robustness with an average F1-Score of 0.850 ± 0.120 and intersection over union score of 0.760 ± 0.150 on the test set. Clinical reliability is assessed with mean absolute error for diameter and length measurements respectively of 1.73 mm, and 2.27 mm. 3D surface reconstruction demonstrated reliability, low processing times, and clinically valid reconstructions. Conclusions: The proposed algorithm can correctly reconstruct pathological vessels. Secondary aorto-iliac pathologies are detected properly for challenging anatomies. To conclude, 3D reconstruction’s application to a digital, patient-specific diagnostic tool is therefore possible. Automatic replicable pipelines ensured usability of the model’s outputs.