Preoperative planning (PP) considering the patient-specific morphology based on medical images is an established process in total hip arthroplasty (THA). Although the high importance of functional parameters for the optimal alignment of the prosthesis is well-known from literature, these parameters are not usually considered in the PP. Prediction models for postoperative functional parameters can be integrated into the PP. However, very little attention has been paid to the accuracy of the prediction. Therefore, the aim of the thesis was to quantify the prediction uncertainties of morpho-functional prediction models for postoperative functional parameters and reveal the implications for a patient-specific PP. The prediction of the postoperative pelvic tilt in standing position was evaluated using multiple linear regression models that included preoperative parameters of a cohort of 196 THA patients. Static musculoskeletal models for the prediction of the hip joint force were adapted by the osseous morphology of the patient and validated using in vivo data of ten THA patients. Selected prediction models of both functional parameters were integrated into the PP process and the necessary reduction of a patient-specific target zone by the prediction uncertainties was investigated. The results show that functional parameters can change significantly after THA and the postoperative values can be predicted preoperatively. The prediction uncertainty of each prediction model has to be considered in the PP process. The prediction uncertainties have a significant impact on the preoperative plan. This highlights the importance of accurate preoperative parameter measurements and further research to improve the prediction of postoperative functional parameters.weiterlesen