Computer-aided shape features extraction and regression models for predicting the ascending aortic aneurysm growth rate

TL;DR

This work tackles the challenge of predicting ascending aortic aneurysm growth by comparing local geometric features with global shape features derived from statistical shape analysis. It introduces a geometry-based pipeline that combines segmentation, centerline-driven local feature extraction, iso-topological mesh morphing, and regression using both PCA and PLS-derived global features to predict growth rates. The study demonstrates that global features, particularly when modeled with PLS, yield more accurate growth predictions than local features, highlighting the value of incorporating shape context into risk assessment. The findings suggest that integrating shape-based biomarkers with biomechanical or hemodynamic data could enhance personalized monitoring and decision-making for AsAA management.

Abstract

Objective: ascending aortic aneurysm growth prediction is still challenging in clinics. In this study, we evaluate and compare the ability of local and global shape features to predict ascending aortic aneurysm growth. Material and methods: 70 patients with aneurysm, for which two 3D acquisitions were available, are included. Following segmentation, three local shape features are computed: (1) the ratio between maximum diameter and length of the ascending aorta centerline, (2) the ratio between the length of external and internal lines on the ascending aorta and (3) the tortuosity of the ascending tract. By exploiting longitudinal data, the aneurysm growth rate is derived. Using radial basis function mesh morphing, iso-topological surface meshes are created. Statistical shape analysis is performed through unsupervised principal component analysis (PCA) and supervised partial least squares (PLS). Two types of global shape features are identified: three PCA-derived and three PLS-based shape modes. Three regression models are set for growth prediction: two based on gaussian support vector machine using local and PCA-derived global shape features; the third is a PLS linear regression model based on the related global shape features. The prediction results are assessed and the aortic shapes most prone to growth are identified. Results: the prediction root mean square error from leave-one-out cross-validation is: 0.112 mm/month, 0.083 mm/month and 0.066 mm/month for local, PCA-based and PLS-derived shape features, respectively. Aneurysms close to the root with a large initial diameter report faster growth. Conclusion: global shape features might provide an important contribution for predicting the aneurysm growth.

Figures (8)

• Figure 1: Pipeline of the full procedure to assess the growth rate from local and global shape features.
• Figure 2: (A) Spheres built around the markers to acquire the grey levels for the thresholding method. (B) Final model derived from the segmentation.
• Figure 3: (A-C) Local shape features for the ascending aorta model; (D) above, the model with the lateral splines on the surface. Below, a section $\Psi_l$ of the aorta for computing the points that identified the splines.
• Figure 4: (A) SSM compactness according to the number of PCA modes used: the symbols indicate the PCA mode for which 80%, 90%, 95% and 99% of the cumulative variance is reached. (B) Generalisation curve when increasing the number of PCA modes. (C) $FS$ for selecting the PCA modes to perform SVM regression. (D) $FS$ for the PLS modes.
• Figure 5: Shape modification due to PCA modes 1, 2 and 6 and PLS modes 1, 2 and 3.