The accuracy of acquisition of an imageless computer-assisted system and its implication for knee arthroplasty

The majority of the current computer-assisted systems (CAS) for knee prosthetic surgery require the acquisition of points using a tracker detected by an infrared camera. Two types of measurements are then essential: angles and distances. The goal of this study was to evaluate the accuracy of the data obtained during computer-assisted surgery using an in vitro protocol. Two models were developed to locate precisely both points (120 acquisitions) and distance measurements (144 acquisitions) and angles (170 acquisitions) with an image-free CAS using an infrared optical camera. For validation, a precise coordinate 3D measurement device was used to assess the accuracy of CAS acquisitions. The points, distances and angles had a mean error respectively of 0.638 mm (0.244 mm to 0.931 mm), 0.355 mm (0.001 mm to 1.338 mm) and 0.39° (0.06° to 0.69°). For all these acquisitions, the mean error was statistically less than 1mm or 1° (p<0.001). By using a 3D measurement system, it was possible to determine the accuracy of the data obtained with the navigation system. The precision assessed at less than 1mm or 1° corresponds with the accuracy needed in knee arthroplasty and with the use of CAS as a measurement tool. Variability in computer-assisted navigation may be more likely related to other aspects of the procedure such as bony landmark acquisition, tolerance of saw blades and cutting jigs than the accuracy of the imageless CAS itself.