Image Rotation Alters Apparent Fibula Length: An Evaluation of Talocrural Angle, Shenton Line, and Dime Sign.

Abstract

Fibula shortening can compromise ankle stability and force transmission, thereby impacting clinical outcomes. Because radiographs depict 3-dimensional anatomy in 2 dimensions, accurate radiographic assessment of fibula length is a commonly encountered clinical challenge. The talocrural angle (TCA), Shenton line, and dime sign are useful parameters of fibula length. Yet, the impact of 3-dimensional limb positioning on these radiographic parameters is not established. Bone models were constructed from CT scans of 30 lower limbs. Fibula length was computationally manipulated, and digitally reconstructed radiographs were generated reflecting 1-degree increments of sagittal and axial plane rotation of each limb for each fibula length condition. The TCA was computationally measured on each image. The presence of an aligned mortise view, intact Shenton line, and intact dime sign was assessed by 2 observers. The mean TCA, which was 78.0 (95% CI ± 1.6) degrees for a true mortise projection with anatomic fibula length, changed by approximately 1 degree per millimeter of fibula length change. On average, 14.7 degrees of caudal rotation obscured 2 mm of fibular shortening by virtue of producing the same TCA as a true mortise view with anatomic fibula length, designated a false positive view. Axial rotation had a comparatively small effect. Observers 1 and 2 were, respectively, 91% and 88% less likely to accurately judge the image alignment of the false positive images compared to true mortise images. Moreover, intraobserver agreement was poor to moderate (mean 0.47, range 0.13-0.59) and interobserver agreement was uniformly poor (mean 0.08, range 0.01-0.20). In our study using digitally reconstructed radiographs from CT scans of 30 limbs, we found that sagittal plane rotation impacts the radiographic appearance of fibula length as measured by the TCA. Limb axial rotation had a comparatively small effect. Further study of human perception of Shenton line and dime sign is needed before the effect of rotation on these parameters can be fully understood. Level IV, case series.