INTRODUCTION Assessing femoral head coverage is a crucial element in acetabular surgery for hip dysplasia. CT has proven to be more accurate, practical and informative than plain radiography at analysing hip geometry. Klaue et al first used a computer-assisted model to indirectly derive representations of femoral head coverage. Jansen et al then described a CT-based method for measuring centre edge angle of Wiberg at 10 rotational increments. Haddad et al used that method to look at dysplastic hips pre- and post-acetabular osteotomy. We present a novel CT-based method that automatically gives an image of the head with the covered area precisely represented. We used this technique to accurately measure femoral head coverage (FHC) in normal hips and in a prospective study of patients with hip dysplasia undergoing peri-acetabular osteotomy. The impact of surgery on acetabular anteversion and inclination was also assessed.

METHODS Using a custom software programme, anatomical landmarks for 25 normal and 26 dysplastic hips were acquired on the 3D reconstructed CT image and used to define the frame of reference. Points were then assigned on the femoral head surface and the superior half of the acetabular rim after aligning the pelvis in the anterior pelvic plane. The programme then automatically produced an image representing the femoral head and its covered part along with the calculated femoral head coverage. To do so, the software represents the femoral head by a best-fit sphere, and the sphere and the acetabular contour are then projected onto a plane in order to calculate the load bearing fraction and area.

RESULTS In the normal hips FHC averaged 73% (SD 4), whereas anteversion and inclination averaged 16° (SD 7°) and 44° (SD 4°) respectively. In the dysplastic group the mean FHC was 50% (SD 6), with a mean anteversion of 19° (SD 10°) and mean inclination of 53° (SD 5°). Peri-acetabular osteotomy has been performed on 16 hips so far, and the FHC for those averaged 66% (SD 5), a mean improvement of 32%. The respective anteversion and inclination post-operatively were 18° (SD 12°) and 40° (SD 8°).

DISCUSSION This is the first study to our knowledge that has used a reliable and practical measurement technique to give an indication of the percent coverage of the femoral head by the acetabulum in normal hips. When this is applied to assessing coverage in surgery to address hip dysplasia it gives a clearer understanding of where the corrected hip stands in relation to a normal hip, and this should allow for better determination of the likely outcome of this type of surgery. The versatility of the method gives it significant attraction for acetabular surgeons and makes it useful not only for studying dysplastic hips but also other hip problems such as acetabular retroversion.