Three-Dimensional Reconstruction Modeling of the Spatial Displacement, Extent and Rotational Orientation of Undisplaced Femoral Neck Fractures

The purpose of this study was to employ a new three-dimensional (3D) reconstruction and modeling method to measure displacement of undisplaced femoral neck fractures (Garden stages I and II). We also aimed to evaluate the effectiveness of the Garden classification for determining the displacement of undisplaced femoral neck fractures. A total of 120 consecutive patients with undisplaced femoral neck fractures were enrolled between 2012 and 2014, including 60 within the Garden I group and 60 within the Garden II group. The displacements of the femoral head center (d1) and the lowest point of the fovea capitis femoris (d2) and rotational displacement of the femoral head (α) in the 3D model were measured with 3D computed tomography reconstruction and modeling. Five observers, trauma surgeons, were asked to found the centers of the femoral heads and the deepest points of the foveae. The intraobserver and inter-observer agreements were calculated using Fleiss' kappa. The inter-observer and intra-observer kappa values were 0.937 and 0.985, respectively. Current method has good reliability. We discovered that many participants in our study had been misclassified by an anterior-posterior radiograph as having an "incomplete" fracture. In incomplete fracture of Garden stage I group, the average displacements d1 and d2 were 3.69 ± 1.77 mm and 14.51 ± 1.91 mm, respectively. The mean α was 4.91° ± 2.49°. For impacted fracture of Garden stage I, significant spatial displacement in the impacted fractures was observed (d1: 6.22 ± 3.36 mm; d2: 10.30 ± 5.73 mm; and α: 17.83° ± 10.72°). Similarly, significant spatial displacement was observed among the Garden stage II group (d1: 7.16 ± 4.58 mm; d2: 12.95 ± 8.25 mm; and α: 18.77° ± 9.10°). There was no significant difference in α, d1, and d2 between impacted fracture and Garden stage II groups (P > 0.05). However, significant differences were found between incomplete fracture and Garden stage II groups (P < 0.05). Our findings suggest that 3D reconstruction and modeling may be a better tool for assessing femoral neck fractures than the Garden classification. Undisplaced femoral neck fractures showed variable degrees of displacement and were not undisplaced, stable fractures. Garden classification for undisplaced femoral neck fractures has certain limitations.