ARTHROSCOPIC FAI SURGERY-PRECISION OF A NAVIGATED REGISTRATION TECHNIQUE

Abstract

Background: Arthroscopic femoral osteoplasties might cause prolonged operative times, restricted intraop-erative overview or insufficient localisation of surgical tools. Computer assisted techniques should improve the precision with an overall accuracy is within 1mm/1°. An automated navigated registration process matching preoperative CT data and intraoperative fluoroscopy, should allow for non-invasive registration for FAI surgery. We evaluated the general precision (I) of the CT and fluoroscopic matching process and (II) the precision of identifying the defined osseous lesions in various anatomical areas.

Material and Methods: Three cadavers (6 hip joints) utilizing a conventional navigation system were used. Before preoperative CT scans, defined osseous lesion (0.5x0.5mm) in the femoral neck, head neck junction, head region were created under fluoroscopic control. Following reference marker fixation, two fluoroscopic images (12 inch c-arm) with 30° angle differences of the hip joint were taken. Automated segmentation including CT-fluoro image fusion by the navigation system enabled a noninvasive registration process Precision of registration process was tested with a straight navigated pointer (1mm tip) trough a lateral arthroscopic portal, during virtual contact to the bone, without arthroscopic control After arthroscopic view was enabled the in vivo distance of pointer tip to bone was measured (I). In vivo real distances between inserted navigated shaver and osseous lesions was done over an anterior hip arthrotomy. Under navigated control, blinded to the situ, placement in the lesions should be done. Distances between shaver tip and osseous lesions were measured with a caliper (II).

Results: The precision for registration (I) was within 0.9mm within the femoral neck (SD 0.24mm; 0.6–1.3mm); 1.2 mm (SD 0.33mm; 0.8–2.0mm) (p> 0.05) for the head neck junction; 2.9 mm (SD 0.57mm; 1.8–3.7mm) for the femoral head (p< 0.001 respectively p< 0.001) Mean offset of the navigated shaver to the lesions (II) was 0.93 mm (SD 0.65mm; 0–2mm). Within the femoral neck a mean accuracy of 0.6mm (SD 0.59mm; 0–1.4mm), the head neck junction 0.8 mm (SD 0.78mm; 0.1–1.5mm), the femoral head 1.3 mm (SD 0.50mm; 0.6–1.7mm) was found (p> 0.05; p> 0.05; p> 0.05).

Conclusion: A combined CT-fluoroscopy matching procedure allows for a reproducible noninvasive registration process for navigated FAI surgery. Precision of the registration process itself is more accurate at the femoral neck and head-neck junction than at the femoral head area. However a navigated identification of osseous lesions was possible within 1mm deviations in all regions.