Application of three-dimensional computerised tomography reconstruction and image processing technology in individual...

Acetabular coverage deficiency displays individual difference among patients with developmental dysplasia of the hip (DDH). Therefore, the correct direction and degree of the acetabular fragment is patient-specific during Bernese periacetabular osteotomy (PAO). This paper introduces a feasible method using 3D computed tomography (CT) and computer image processing technology for customised surgical planning.

CT data of 96 hips in 60 DDH patients (male 15, female 45; average age/range 30 ± 8/14-49 years) and 53 normal hips (male 13, female 37; average age/range 52 ± 13/16-69 years) were reconstructed using commercially available software Mimics and Imageware. Geometric parameters of each hip were measured in relation to the anterior pelvic plane after correcting for pelvic tilt and rotation. Deficiency types and degrees of acetabular dysplasia in patients with DDH were determined by comparison with normal hips, and improvement in femoral-head coverage was analysed again after virtual PAO. A customised surgery programme for each DDH patient was designed and provided the reference for the actual operation.

We produced a 3D pelvic model using image processing software, doing precise measurement and with close approximation to the actual PAO. Lateral centre-edge angle (LCEA), anterior centre-edge angle (ACEA), acetabular anteversion angle (AAVA), anterior acetabular sector angle (AASA) and posterior acetabular sector angle (PASA) of normal hips in the control group were 35.128 ± 6.337, 57.052 ± 6.853, 19.215 ± 5.504, 61.537 ± 7.291 and 99.434 ± 8.372°, respectively. Angles of hips with DDH before surgery were 11.46 ± 11.19, 35.79 ± 13.75, 22.77 ± 6.13, 43.58 ± 9.15 and 88.46 ± 8.24, which were corrected to 33.81 ± 2.36, 55.38 ± 2.09, 20.16 ± 2.18, 58.29 ± 7.60, and 4.71 ± 7.75°, respectively, after surgery. After virtual Bernese PAO, LCEA, ACEA, AAVA, AASA and PASA were corrected significantly (p < 0.01). There was no statistically significant differences between LCEA, ACEA and AAVA after virtual Bernese PAO and normal hips (p = 0.06, p = 0.23, p = 0.06°, respectively). AASA improved significantly (p = 0.002) post-operatively at the cost of reducing posterior coverage represented by PASA, which is significantly smaller than in normal and pre-operative hips of DDH patients (p < 0.01).

The geometric feature of the pelvis for patients with DDH can be assessed comprehensively by using 3D-CT reconstruction and image processing technology. Based on this method, surgeons can design individualised treatment scheme and improve the effect of PAO.