A Region-Based Statistical Shape Modeling on the First Trapezoid-Metacarpal.
Lin FuKaczmarek ŁukaszCi JiangYaodong GuPublished in: Applied bionics and biomechanics (2023)
A statistical shape model (SSM) based analysis was conducted in our study. We indicated the morphological differences of the first trapezoid-metacarpal (TMC) bone between the region-based statistical shape model (rSSM) and the nonregion-based statistical shape model (nrSSM). We aimed to characterize more specific and accurate variation results to detect and prevent osteoarthritis in an early way. CT image data of TMC from 31 healthy volunteers were simulated with 3D models. A training set of models was analyzed with principal component analysis, with both then- rSSM and rSSM. The region was identified automatically with Gaussian curvature analysis. We found four dominant shape variations from both nrSSM and rSSM. The rSSM showed more variations with fewer components compared with nrSSM. The first principal component revealed the size variation as the biggest component. A significant difference was presented in the second and the fourth principal component from rSSM, showing the torsion of the distal head of the first metacarpal which may increase the risk of osteoarthritis.
Keyphrases
- rheumatoid arthritis
- computed tomography
- knee osteoarthritis
- magnetic resonance imaging
- high resolution
- electronic health record
- bone mineral density
- single cell
- minimally invasive
- machine learning
- mass spectrometry
- big data
- data analysis
- postmenopausal women
- body composition
- artificial intelligence
- pet ct
- soft tissue