Proximal femoral medullary canal diameters in Indians: correlation between anatomic, radiographic, and computed tomographic measurements.

PURPOSE To compare anatomic, radiographic, and computed tomographic (CT) measurements of the proximal femur in an Indian population. METHODS 26 left and 24 right dried cadaveric femurs were obtained. Each femur was divided into segments at 10 cross-sectional levels from proximal to distal. At each level, anteroposterior (AP), lateral, 45-degree internal oblique, and 45-degree external oblique diameters of the medullary canal were measured using radiography and CT. To minimise magnification, the femur was placed over the film/board. The neck shaft angle, femoral head offset, neck length, and femoral head diameter were also measured. For anatomic measurements, the cross-sections of the medullary canal at levels 3 to 10 were measured using a calliper. Anatomic measurements were compared with radiographic and CT measurements, and the distortion calculated. Correlation between the 3 modalities was calculated using the Pearson correlation coefficient. Accuracy index of the 2 diagnostic modalities was measured, based on the differences between anatomic and radiographic/CT measurements. Lower accuracy indices indicated higher accuracy. RESULTS On AP radiographs of all femurs, the mean femoral head diameter was 45 mm, the mean neck shaft angle was 132 degrees, the mean femoral head offset was 42 mm, and the mean neck length was 63 mm. With regard to medullary canal diameters, the radiographic and CT measurements were smaller than anatomic measurements at proximal levels, but similar at distal levels. Variations between femurs and between the 4 diameters at the same level were greater at proximal levels. CONCLUSIONS Combined use of radiography and CT is recommended for preoperative assessment of implant size, particularly in uncemented arthroplasty where an optimal fit is essential for biological fixation.