Unique shape variations of hind and midfoot bones in flatfoot subjects—A statistical shape modeling approach

Abstract

Flatfoot deformity is a prevalent hind‐ and midfoot disorder. Given its complexity, single‐plane radiological measurements omit case‐specific joint interaction and bone shape variations. Three‐dimensional medical imaging assessment using statistical shape models provides a complete approach in characterizing bone shape variations unique to flatfoot condition. This study used statistical shape models to define specific bone shape variations of the subtalar, talonavicular, and calcaneocuboid joints that characterize flatfoot deformity, that differentiate them from healthy controls. Bones of the aforementioned joints were segmented from computed tomography scans of 40 feet. The three‐dimensional hindfoot alignment angle categorized the population into 18 flatfoot subjects (≥7° valgus) and 22 controls. Statistical shape models for each joint were defined using the entire study cohort. For each joint, an average weighted shape parameter was calculated for each mode of variation, and then compared between flatfoot and controls. Significance was set at p < 0.05, with values between 0.05 ≤ p < 0.1 considered trending towards significance. The flatfoot population showed a more adducted talar head, inferiorly inclined talar neck, and posteriorly orientated medial subtalar articulation compare to controls, coupled with more navicular eversion, shallower navicular cup, and more prominent navicular tuberosity. The calcaneocuboid joint presented trends of a more adducted calcaneus, more abducted cuboid, narrower calcaneal roof, and less prominent cuboid beak compared to controls. Statistical shape model analysis identified unique shape variations which may enhance understanding and computer‐aided models of the intricacies of flatfoot, leading to better diagnosis and, ultimately, surgical treatment.

The relationship of three‐dimensional foot morphology to clinical assessments and postural stability in adolescent male footballers

“…Transformed STL files were processed through an established SSMs framework [29, 30], to generate a SSM of the foot for the entire population: First, meshes from one subject were non‐rigidly fit to all the remaining segmentations to ensure mesh and node correspondence. Second, all meshes were rigidly aligned to remove global rotational and translational variations.…”