OCLC: 988603751
PROQUEST: 1871371992
Background
Correct setting of femoral anteversion (FA), femoral offset (FO), and vertical offset (VO) is important to restoring function in total hip arthroplasty (THA). When poorly set, FA, FO, and VO have been associated with poor hip functions. Therefore most THA procedures are preceded by preoperative planning during which a surgeon explores different femoral component designs, sizes, and configurations to obtain the combination that provides the best setting of FA, FO, and VO and best fitting of the component in the medullary canal. Although there is no consensus in setting acceptable limits for these variables, FA set within ± 5° and FO and VO set within ± 5 mm of that of the native limb are clinically desirable. Setting of these clinical variables depends on the placement of the femoral component in the femur, and for cementless THA, this is dictated by the position and orientation of the femoral cavity. Therefore, machining a femoral cavity with position and orientation accurate to the planned cavity is important.
While a surgical robotic system developed for total hip arthroplasty (THA) offers advantages over manually-performed total hip arthroplasty procedures, there are various sources of inaccuracy which can lead a robotic system to machine a cavity with different position and orientation than that planned. This in turn can translate into a femoral component placement that incorrectly sets the FA, FO, and VO.
At present, only one study reported position and orientation errors of roboticallymachined femoral cavities in six degrees of freedom. However, the results were of an older generation robotic system using pin-based registration. Accuracy for a robotic system using the new surface-based registration is unknown. Also, unknown is the accuracy in setting FA, FO, and VO when a femoral component is placed in a robotically-machined cavity. Accordingly, the objectives of this thesis were to:
Methods
After creating preoperative plans, robot-assisted THAs were performed on twelve cadaveric specimens. 3D models of machined cavities were created. The position and orientation of the machined cavities were compared to those of the planned cavities by best-fitting. Position and orientation of the machined cavity relative to the planned cavity were computed as errors in six degrees of freedom (medial/lateral, anterior/posterior, and proximal distal translations; flexion/extension, varus/valgus, and internal/external rotations) allowing the bias (mean error) and precision (standard deviation of error) to be determined for each degree of freedom.
The planned femoral components were manually seated into the robotically-machined cavities. The achieved and planned femoral component placements were measured, and the error between achieved and planned component placements was computed allowing the bias and precision to be determined for FA, FO, and VO. The percent of the population outside of clinically desirable limits of ± 5° and ± 5 mm was determined for each variable.
Results
Bias in the machined cavity occurred for anterior and proximal translations and extension rotation. The precision in the machined cavity was less than 1.5 mm for all translational degrees of freedom and less than 1.2° for rotational degrees of freedom.
When differences between the achieved and planned femoral component placements were determined, bias of 1.5° and 2.7 mm occurred for FA and VO but not FO. The precision -vwas less than 2° for FA and less than 2 mm for FO and VO. The percent of population outside of the limits of ± 5° and ± 5 mm was 0.4% for FA, 0.0% for FO, and 5.0% for VO.
Discussion
Using robot-assisted THA, biases in three of six degrees of freedom of femoral cavities machined by a new generation robotic system were statistically significant. However, the position and orientation errors in the robotically machined femoral cavity were such that, when placing the planned femoral components into the cavities, FA, FO, and VO were consistently within the clinically desirable limits of ± 5° and ± 5 mm for at least 95% of the population. Hence the use of an active robot may decrease the incidence of adverse clinical outcomes.