Abstract
Adolescent Idiopathic Scoliosis (AIS) characterized by the 3D spine deformity affects about 4% schoolchildren worldwide. One of the prominent theories of the etiopathogenesis of AIS was proposed to be the poor postural balance control due to the impaired vestibular function. Thus, the morphometry of the vestibular system (VS) is of great importance for studying AIS. The VS is a genus-3 structure situated in the inner ear and consists of three semicircular canals lying perpendicular to each other. The high-genus topology of the surface poses great challenge for shape analysis. In this work, we propose an effective method to analyze shapes of high-genus surfaces by considering their geodesic spectra. The key is to compute the canonical hyperbolic geodesic loops of the surface, using the Ricci flow method. The Fuchsian group generators are then computed which can be used to determine the geodesic spectra. The geodesic spectra effectively measure shape differences between high-genus surfaces up to the hyperbolic isometry. We applied the proposed algorithm to the VS of 12 normal and 15 AIS subjects. Experimental results show the effectiveness of our algorithm and reveal statistical shape difference in the VS between right-thoracic AIS and normal subjects.
Introduction
Adolescent Idiopathic Scoliosis (AIS) is a complex three-dimensional spinal deformity that affects approximately 4% of children globally. Despite its prevalence, the underlying cause remains elusive. A leading hypothesis suggests that AIS may be linked to impaired postural balance, which is primarily regulated by the vestibular system (VS) in the inner ear.
The VS presents a significant anatomical challenge for traditional shape analysis. It is a "genus-3" surface, topologically equivalent to a triple-holed torus, comprising three semicircular canals. Conventional tools designed for "genus-0" shapes (like the brain or heart) are inadequate for these high-genus structures. This research addresses the need for a mathematically rigorous framework to compare these complex surfaces and investigate potential structural abnormalities correlating with spinal curvature.
Methods
The researchers developed a pipeline rooted in conformal geometry and hyperbolic metrics to analyze the VS:
- Ricci Flow: The VS surface mesh is deformed into a canonical hyperbolic space using the Ricci flow method. This process assigns a hyperbolic metric to the surface, mapping the complex 3D anatomy onto a Poincaré disk.
- Fuchsian Group Generators: In the hyperbolic domain, the fundamental symmetries of the surface—the Fuchsian group—are computed.
- Geodesic Spectra: For every homotopy class of loops on the surface, there exists a unique shortest path known as a geodesic. The lengths of these loops constitute the "geodesic spectrum."
- Shape Fingerprinting: These lengths are invariant under isometry, meaning they remain constant regardless of the shape's orientation or position. Consequently, they serve as a unique geometric "fingerprint" for the vestibular canals.
Theoretical Framework: Shape Fingerprinting
Instead of simplifying the VS into a sphere, this method embraces its genus-3 topology. The geodesic spectrum naturally captures the dimensions of the three semicircular canals:
- $a_i$ values correspond to the length of the canals.
- $b_i$ values correspond to the thickness or circumference of the canals.
By utilizing hyperbolic geometry, the researchers identify stable descriptors that are mathematically guaranteed to represent the surface's intrinsic structure, effectively filtering out noise from different poses or imaging resolutions.
Datasets
The study utilized a clinical dataset of 27 subjects:
- AIS Group: 15 girls with right-thoracic AIS (mean age 15, mean Cobb’s angle 27.27°).
- Control Group: 12 age-matched healthy girls.
- Imaging: High-resolution T2-weighted MRI scans (1.5T) were used to segment the left-side vestibular system, generating triangle meshes with approximately 50,000 faces each.
Results
Statistical analysis (t-tests) revealed significant morphological differences between the AIS and control groups:
- Morphological Trends: AIS patients generally exhibited longer ($a_i$) and thinner ($b_i$) vestibular canals.
- Lateral Canal Significance: The lateral canal showed the most pronounced differences ($P < 0.05$). The ratio of length to thickness ($a_1/b_1$) was a particularly strong differentiator ($P = 0.0338$).
- Global Thinning: The sum of thickness parameters ($b_1 + b_2 + b_3$) was highly significant ($P = 0.0040$), suggesting a systemic "slenderizing" of the vestibular system in AIS patients.
- Computational Efficiency: The automated pipeline is highly efficient, processing each subject in approximately 25 seconds on standard hardware.
Discussion and Future Work
While these findings provide a potential structural link to AIS, several areas for future research remain:
- Sample Size: A larger cohort is necessary to validate these findings as a reliable clinical biomarker.
- Bilateral Analysis: This study focused on the left-side VS; future work should investigate bilateral asymmetry, which is common in balance disorders.
- Causality: The study establishes correlation but not causation. It remains unclear if vestibular changes cause AIS or result from compensatory mechanisms for spinal curvature.
- Clinical Application: Future efforts could involve developing predictive classifiers using machine learning on the geodesic spectra for early AIS detection.
Summary updated