Do hip resurfacing and short hip stem arthroplasties differ from conventional hip stem replacement regarding impingement‐free range of motion?

Total hip joint replacement (THR) is clinically well‐established. In this context, the resulting range of motion (ROM) is crucial for patient satisfaction when performing joint movements. However, the ROM for THR with different bone preserving strategies (short hip stem and hip resurfacing) raises the question of whether the ROM is comparable with conventional hip stems. Therefore, this computer‐based study aimed to investigate the ROM and type of impingement for different implant systems. An established framework with computer‐aided design 3D models based on magnetic resonance imaging data of 19 patients with hip osteoarthritis was used to analyse the ROM for three different implant systems (conventional hip stem vs. short hip stem vs. hip resurfacing) during typical joint movements. Our results revealed that all three designs led to mean maximum flexion higher than 110°. However, hip resurfacing showed less ROM (−5% against conventional and −6% against short hip stem). No significant differences were observed between the conventional and short hip stem during maximum flexion and internal rotation. Contrarily, a significant difference was detected between the conventional hip stem and hip resurfacing during internal rotation (p = 0.003). The ROM of the hip resurfacing was lower than the conventional and short hip stem during all three movements. Furthermore, hip resurfacing shifted the impingement type to implant‐to‐bone impingement compared with the other implant designs. The calculated ROMs of the implant systems achieved physiological levels during maximum flexion and internal rotation. However, bone impingement was more likely during internal rotation with increasing bone preservation. Despite the larger head diameter of hip resurfacing, the ROM examined was substantially lower than that of conventional and short hip stem.


| INTRODUCTION
Total hip joint replacement (THR) remains the gold standard therapy option for end-stage hip osteoarthritis and is currently performed more than 240,000 times in Germany alone. 1 In this regard, the incidences are expected to increase with demographic aging and life expectancy. 2THR using conventional hip stems has proven clinically and functionally successful.Despite these successes, the treatment of increasingly younger, more active patients presents the challenge of preserving the proximal femoral bone stock for later revisions 3 to enable cementless revision with a conventional hip stem.Therefore, in younger patients with adequate bone quality, 4,5 short hip stem prostheses, which aim to preserve most of the proximal femur, [4][5][6][7] and hip resurfacing implants, which preserve the femoral neck completely, 8 have been introduced into clinical use. 7,9Short hip stem prostheses theoretically offer the advantages over the conventional hip stem of providing a more physiologic loading of the proximal femur 5,7,[10][11][12][13] and reducing the risk of stress-shielding. 7,10Alternatively, hip resurfacing arthroplasty (HRA) is known for being bone preserving 9,14,15 and for retaining physiological joint geometry as well as physiologic reconstruction of the joint mechanics. 9,16In this context, the resulting low bone loss and, thus the good revision possibilities due to sufficient remaining bone material could be an indication for considering a short hip stem and HRA, respectively, especially in a younger patient group with hip osteoarthritis. 15,17wever, different implant geometries result in significant variation of the postoperative shape of the proximal femur: Fullsize neck resection and implantation of a conventional hip stem lead to substantial reduction of bone around a relatively thin implant taper.Due to a more proximal femoral neck cut in most short stem implants, the amount of retained metaphyseal bone is higher when compared with standard stems.In HRA, the femoral neck is completely preserved, resulting in a substantially larger neck diameter than in conventional and short stem implants.
The different implant and bone resection philosophies may impact the postoperative range of motion (ROM) and potential impingement, which raises the question of the extent to which short hip stems and HRA, respectively, allow sufficient ROM of the hip joint compared with conventional hip implants.9][20][21][22][23] Previous studies on the ROM of hip implants have indicated that dislocation and impingement of the femoral neck are influenced by implant positioning 24 and the implant design. 8,25ere are only a few studies on ROM and impingement behavior, which compare HRA and short hip stem systems with conventional standard stems within different patients. 26It is unclear, however, whether the impingement behavior and the ROM for short hip stems and hip resurfacing is different compared with a conventional hip stem when applied at the same patient.Moreover, knowledge of the clinically relevant parameter of ROM in dependence on the design principle is limited.Therefore, we present a noninvasive computational study based on MRI data of 19 patients with hip osteoarthritis to investigate the ROM between hip resurfacing and a short hip stem design compared with a conventional stem.

| METHODS
Using a previously described computational framework based on medical imaging and computer aided design (CAD) data, 27 we analyzed the influence of three different implant designs on postoperative ROM according to a preoperatively planned situation and a postoperatively reconstructed situation during three clinically relevant motions.
For the investigation of the ROM, we used three different implant systems (Figure 1): an HRA system (Durom; Zimmer Biomet), a conventional hip stem (BiContact; Aesculap AG), and a short hip stem (Metha; Aesculap AG), each in combination with a conventional press-fit acetabular cup (Plasmacup; Aesculap AG) and a ceramic-onceramic bearing with 28-36 mm diameter BIOLOX-delta ceramic heads (CeramTec).We included 19 patients (8 females, 11 males) with a mean age of 52.1 ± 7.2 years in our study.According to their preference, five of them have been scheduled for HRA.The remaining 14 patients were randomized for either a conventional or a short hip stem.Six of these patients received treatment with the conventional stem, and eight patients were treated using the short hip stem.It has to be noted that in the resurfacing group, one patient received a different implant, which cannot be used for comparing the preoperatively planned situation with the postoperatively F I G U R E 1 Depiction of the different total hip implant systems used in this study: conventional hip system (A), short hip stem system (B), and hip resurfacing arthroplasty (C).
reconstructed situation but for comparing the implant systems in the preoperatively planned situation.The study design was approved by the local ethics committee at the University Hospital Carl Gustav Carus, Technical University of Dresden (EC No. 175052011).Written informed consent was obtained from all patients before enrolling.
Based on noninvasive MRI data, our CAD framework was used to analyze the ROM and the type of impingement for the conventional hip stem according to a preoperatively planned situation and a postoperatively reconstructed situation during three clinically relevant movements: [28][29][30][31] maximum flexion starting from neutral (MaxFlex), internal rotation at 90°flexion (IR@90Flex) and external rotation at 90°flexion (ER@90Flex).Briefly, we compared the ROM between the preoperatively planned situation and the preoperatively reconstructed situation for six patients who received the conventional hip stem.Additionally, we analyzed the influence of implantspecific parameters (e.g., size of the prosthetic head and acetabular cup) of the conventional hip stem on the ROM for the preoperatively planned situation in 19 patients.
The workflow of this study is depicted in Figure 2. Firstly, we derived preoperative bone geometries using segmentation and reconstruction techniques based on the coordinate systems according to our previous studies. 8,27Secondly, we virtually implanted the three implant designs resembling two situations for each patient: a preoperatively planned situation based on preoperative MRI scans as well as radiographs and a preoperatively reconstructed situation based on postoperative radiographs.Each virtual implantation was conducted under the supervision of an experienced surgeon and according to the surgical manual of the implant manufacturer.Finally, the ROM was determined using a kinematic analysis provided by the CAD software Creo Parametric (v.2.0; Parametric Technology Corporation, Boston, MA, USA) for three clinically relevant joint movements.During the movement analysis, impingement was detected between the femoral configuration and pelvic configuration with a collision detection routine provided by Creo Parametric.We used the previously described workflow to compare the ROM between preoperatively planned and postoperatively reconstructed situations for the subjects treated with the respective implant designs (Figure 2).Additionally, we compared the ROM for the respective implant designs in case of the preoperatively planned situation for all 19 patients.Impingement for the three movements was distinguished between three types: bone-to-bone (BBI), implantto-implant (II), and implant-to-bone (IBI).
Virtual implantation was conducted for the conventional hip stem as previously described for preoperatively planned and postoperatively reconstructed situation. 27Briefly, resection was performed as prescribed in the surgical manual resembling the principle of hip step osteotomy and an osteotomy plane angle of 55°r elative to the femoral shaft axis.The hip stem was aligned parallel to the femoral shaft and aimed to be well-centered in the medullary femoral canal.Femoral antetorsion was determined using the method of Dunlap et al. 32 The acetabular cup was aligned with 45°inclination and 15°anteversion relative to the anatomic planes of the pelvic coordinate system.We also aimed for the best possible reconstruction of the preoperative center of rotation of the femur and acetabulum.Additionally, the postoperatively reconstructed situation was generated using postoperative anterior-posterior radiographs of the patients.The stem was implanted virtually analogous to the procedure for the preoperatively planned situation, using the postoperative radiographs to derive the implant orientation with regard to our previous study. 27Orientation of the acetabular cup was derived according to Murray et al. 33 and Liaw et al. 34 The bone resection necessary for virtually implanting the short hip stem in case of the preoperatively planned situation was performed as prescribed in the surgical manual and according to preoperative planning.Virtual bone resections were performed with an osteotomy plane taken from preoperative planning.The stem was inserted concentrically into the resection plane according to surgical instructions.Attention was paid to the integrity of the cortical outer contour of the remaining femoral neck in the metaphyseal area and the guidance of the prosthesis through the femoral neck.Additionally, we aimed to mimic the alignment of the stem in the preoperative planning based on the preoperative radiograph.The acetabular cup was implanted analogously to the conventional hip stem procedure.
According to the postoperative radiograph, virtual implantation was performed analogously to the procedure described for the preoperatively planned situation.Accordingly, the osteotomy plane was taken from postoperative radiographs and the acetabular cup orientation was derived from Murray et al. 33 and Liaw et al. 34 According to preoperative planning, the femoral component of the HRA was performed by aligning the implant centerline with the femoral neck axis, choosing the implant size as small as possible without affecting the femoral neck.We also aimed to restore the preoperative center of rotation.Additionally, the bone was virtually resected according to the surgical manual provided by the manufacturer.The component was moved parallel to the femoral neck axis for the highly deformed femoral head bone to avoid excessive resection of the bone.The acetabular component was virtually implanted as described for the conventional hip stem.Additionally, the head-toneck ratios were determined from the mean diameters using a bestfit cylinder in the software Geomagic Studio (v.2013; 3D Systems, Rock Hill, SC, USA).
ROM and impingement type were investigated for the respective patients depending on the implant system at the movements MaxFlex, IR@90°Flex, and ER@90°Flex.The parameter ROM is shown as a boxplot with a median to illustrate the distribution within the studied groups.In addition, the mean values with standard deviation are reported to compare the individual groups.Only the mean values with standard deviation were presented for the parameter impingement type.All analyses of the statistical significance were performed using GraphPad Prism 9.2 (GraphPad Software, La Jolla, CA, USA), and a p ≤ 0.05 was considered as statistically significant.First, the t test for connected samples was used to test whether statistically significant differences exist between preoperatively planned situation and postoperatively reconstructed situation for patients who received conventional hip stem (n = 6), patients who received short hip stem (n = 8), and patients who were F I G U R E 2 Workflow of the previously described CAD method 27 based on patient data, e.g.magnetic resonance imaging (MRI) scan, adapted to evaluate the influence of the three different design principles conventional hip stem, short hip stem, and hip resurfacing arthroplasty (HRA) on the resulting range of motion (ROM).treated with HRA (n = 4).Further, the ROM was determined for all 19 patients of each planned implant design according to the preoperatively planned situation during the movement MaxFlex, IR@90°Flex, and ER@90°Flex.The observed ROM of each implant design was evaluated with regard to the standard normal distribution using the Shapiro-Wilk test.Subsequently, a one-way analysis of variance for repeated measures followed by a Bonferroni post hoc test was performed to examine statistically significant differences in the ROM between the implant designs for each movement.The comparison between the different movements was neglected.

| RESULTS
The maximum ROM until impingement was evaluated for all patients depending on the preoperatively planned and postoperatively reconstructed situation.Concerning the postoperatively reconstructed situation, the six patients who received a conventional hip stem, the eight patients who received a short hip stem, and the four patients who received an HRA were compared with the respective preoperatively planned situation (Figure 2).
Regarding the conventional hip stem, there was no significant difference in ROM between the preoperatively planned situation and postoperatively reconstructed situation (Figure 3A) during MaxFlex (p = 0.671) and IR@90Flex (p = 0.858), while there was a significant difference during ER@90Flex (p = 0.036).Similarly, there was also no change between the preoperatively planned situation and postoperatively reconstructed situation during MaxFlex (p = 0.719), IR@90Flex (p = 0.855), and ER@90Flex (p = 0.823) for the patients receiving the short hip stem (Figure 3B).For the patients who received HRA, no significant difference was detected between the preoperatively planned situation and postoperatively reconstructed situation during MaxFlex (p = 0.097), IR@90Flex (p = 0.456), and ER@90Flex (p = 0.267).However, the ROM was decreased for the postoperatively reconstructed situation, for example, in the case of MaxFlex, the preoperatively planned situation showed a mean ROM of 119.7 ± 11.0°.In contrast, the postoperatively reconstructed situation showed a mean ROM of 106.7 ± 20.2°(Figure 3C).
To further evaluate the comparison between preoperatively planned and postoperatively reconstructed situation, the maximum ROM was divided into the impingement types (Figure 4).
For the conventional hip stem (Figure 4A), the predominant type of impingement for the preoperatively planned situation was implant-toimplant impingement during MaxFlex (n = 4), IR@90Flex (n = 6), and ER@90Flex (n = 6) followed by bone impingement during MaxFlex (n = 2).This was also similar to the postoperatively reconstructed situation.Greater angles until impingement were detected for cases where ROM was limited by implant impingement.In the case of bone impingement, a reduction in ROM for MaxFlex from 143.1 ± 2.7°( preoperatively planned situation, BBI) to 140.4 ± 5.4°(postoperatively reconstructed situation, BBI) was determined.
Regarding the short hip stem (Figure 4B), the most likely type of impingement in the preoperatively planned situation was bone-toimplant impingement for MaxFlex (n = 7), IR@90Flex (n = 7), and F I G U R E 3 Box plots for the comparison of the range of motion (ROM) using implantation according to preoperatively planned situation (pps) and reconstructed postoperative situation (prs) for (A): conventional hip stem (n = 6 patients), (B): short hip stem (n = 8 patients) and (C): HRA (n = 4 patients) during maximum flexion (MaxFlex), internal rotation at 90°of flexion (IR@90Flex), and external rotation at 90°flexion (ER@90Flex).Overall, the pps group showed a good agreement with the prs group during the three movements and implant designs.Note that the mean of ROM values is depicted as diamonds.For pps and prs IR@90Flex and ER@90Flex could not be achieved for one patient with HRA as the MaxFlex was less than 90°.Due to this, only n = 3 was used for the comparison between pps and prs for IR@90Flex and ER@90Flex.
Regarding HRA (Figure 4C), a good agreement between the preoperatively planned situation and the postoperatively reconstructed situation was detected in the case of the frequency of each impingement type.The most likely type of impingement for preoperatively planned situation was implant-to-bone impingement for MaxFlex (n = 4), IR@90Flex (n = 3), and ER@90Flex (n = 2).The analysis of the postoperatively reconstructed situation showed a reduction of the maximum ROM during MaxFlex from 119.7 ± 11.0°t o 106.7 ± 20.2°.In comparison, IR@90Flex decreased from 19.1 ± 7°t o 11 ± 5.5°when implant-to-bone impingement occurred.
The maximum ROM until impingement was evaluated for the preoperatively planned situation (n = 19 patients) and the three implant designs (Figure 5).During MaxFlex, there was no significant difference in ROM between the different implant designs.All designs led to a mean MaxFlex higher than 110°.It is noteworthy that the HRA group showed a lower ROM (−5% against conventional and −6% against short hip stem).
Regarding IR@90Flex, there was no significant difference in ROM between conventional and short hip stem (p = 1.000).However, we detected a significant difference between the conventional hip stem and HRA (p = 0.003), where the resurfacing group yielded in lower ROM.Furthermore, there was no significant difference between the short hip stem and HRA (p = 0.089).In the case of the ER@90Flex, there was no significant difference between the three implant designs.
When comparing the short hip stem with the conventional hip stem, there was a slight increase in implant-to-bone impingement during MaxFlex and ER@90Flex (Figure 6A).However, the mean ROM until impingement between conventional and short hip stem is in the same range, for example, the mean ROM during IR@90Flex ranges from 30.9°to 43.3°in case of the conventional hip stem, and 32.6°to 41.4°in the case of the short hip stem (Figure 6A,B).
Comparing HRA with the conventional stem, there were considerable differences in the mean ROM and the type of impingement (Figure 6C).Generally, HRA shifted mostly the impingement type to implant-to-bone impingement (n = 48 during the three movements), whereas only n = 1 patient during the three movements was detected for the conventional hip stem, and n = 8 patients were detected for the short hip stem.
F I G U R E 5 Box plots for the comparison of the range of motion (ROM) using implantation according to the preoperatively planned situation for n = 19 patients for conventional hip stem (CHS), short hip stem (SHS), and hip resurfacing arthroplasty (HRA) during maximum flexion (MaxFlex), internal rotation at 90°of flexion (IR@90Flex), and external rotation at 90°flexion (ER@90Flex).Note that the mean of ROM values is depicted as diamonds.Note that IR@90Flex and ER@90Flex could not be achieved for one patient as the MaxFlex was less than 90°with the HRA.
F I G U R E 6 Quantity N of impingement type during maximum flexion (MaxFlex), internal rotation at 90°of flexion (IR@90Flex), and external rotation at 90°flexion (ER@90Flex) for the three implant designs: conventional hip stem (A), short hip stem (B), and hip resurfacing arthroplasty (HRA) (C).BBI, bone-to-bone-impingement; IBI, implant-to-bone-impingement; II, implant-to-implant-impingement.Note that the impingement for IR@90Flex and ER@90Flex could not be achieved for one patient with HRA as the MaxFlex was less than 90°.
HRA uses the largest heads of the femoral component compared with the conventional and short hip stem.Despite the larger head diameter, this did not lead to higher ROM during the three movements compared with the other implant systems (Figure 7).A further increase of the head diameter from the middle to the largest size allowed more flexion for the conventional hip stem.In contrast, no considerable increase was detectable for the other two designs.

| DISCUSSION
Total hip replacement is a highly successful surgical procedure and results in pain-free mobility of the affected joint in most patients. 35 the surgery is performed increasingly in younger patients, bone preserving implants have been developed and are characterized by reduced metaphyseal bone resection (short hip stems) or even avoidance of neck resection (hip resurfacing).Nevertheless, implants with preserved metaphyseal bone and neck of the femur may lead to earlier impingement than conventional hip stems.As impingement can lead to painful restriction of ROM, component loosening, implant damage or joint instability, 22,23,[36][37][38][39] it is necessary to balance the potential for bone sparing against the risk of a mechanical conflict between the proximal femur and the acetabulum.
According to our computer-based framework, 27 the ROM of three clinically used implant systems, i.e., a conventional hip stem, a short hip stem, and an HRA, which promote different levels of preservation of the proximal femur, were compared in the present study based on preoperatively planned situation and were reconstructed according to the postoperative situation.Furthermore, all three implant systems were preoperatively planned for all patients.Subsequently, the ROM was compared during the movements: maximum flexion, internal rotation at 90°flexion, and external rotation at 90°flexion.
Computer-based studies of ROM determination using simulation models have some inherent limitations.First, the reconstruction of bone geometry from MRI scans should be mentioned.Here, due to the resulting triangulated virtual surface, deviations may occur in the reconstruction process, which can affect the detection of the impingement type in the final model.2][43] In contrast, physiological values of ROM were determined during maximum flexion and internal rotation at 90°f lexion.2][43][44] In our study, all three implant systems were templated in the preoperatively planned situation for all 19 patients with regard to restoration of the individual anatomy.Additionally, this leads to the fact that no systematic variation of the implant parameters CCD angle, implant size, stem neck length, stem neck diameter, and head diameter was performed in a parameter study.Despite this, manual positioning of the femoral components is an influencing factor on ROM 24,45 due to possible inaccurate restoration of the CCD angle because every degree deviation from the optimum CCD angle should be compensated by an adaption of the cup anteversion of 2°and cup inclination of 0.45°to preserve the targeted ROM. 45The head-to-neck ratio could be determined for the conventional hip stem and short hip stem based on the implant neck diameter.Contrarily, the minimum femoral neck diameter for HRA could only be determined as the F I G U R E 7 Range of motion as a function of the prosthetic head diameter during maximum flexion (MaxFlex), internal rotation at 90°of flexion (IR@90Flex), and external rotation at 90°flexion (ER@90Flex) for the three implant designs: conventional hip stem (A), short hip stem (B), and hip resurfacing arthroplasty (HRA) (C).*Note that for MaxFlex n = 4 prosthetic heads of size 42-48 mm could be used for the determination of the ROM depending on the prosthetic head diameter.For IR@90Flex and ER@90Flex, one patient had to be excluded because 90°flexion could not be achieved.mean diameter.This can lead to the assumption of larger neck diameters depending on the position of the native femoral neck.The pelvic geometries generated by MRI imaging were acquired in the supine position. 27Therefore, neither the pelvic tilt during MRI nor the supine position on the operating table reflects the functional pelvic tilt in standing posture. 46,47An average difference in the posterior pelvic tilt of 2°-7°between the preoperative supine position and standing position has been reported in the literature. 48 this context, the pelvic tilt influences the impingement risk less than different neck-shaft-angles or larger head diameters. 25,49wever, new methods to computationally consider the pelvic tilt in standing and sitting should be included, as shown by Tang et al. 47 The comparison between the preoperatively planned and postoperatively reconstructed situation showed good agreement of the resulting impingement types for all three implant systems at maximum flexion and internal rotation at 90°flexion.Conversely, an increase in bone-to-bone impingement was observed during external rotation at 90°flexion.All three implant systems achieved physiological-like levels of ROM for flexion and internal rotation, which agree with the literature. 8,28,42,50,51Furthermore, our simulation results are consistent with previous clinical measurements on patients, [52][53][54] which further underlines the reliability of the modeling approach, as described in our previous study. 276][57] For external rotation at 90°flexion, ROM was overestimated by more than 60°in all three implant systems compared with clinical expectation. 28,31,58This observation was attributed to the missing active and passive parts of the soft tissue structures in the literature, 30,44,59 which agrees with the presented study.Thus, the ROM results for external rotation at 90°flexion should not be considered for direct comparison with clinically determined ROM. 27e ROM differences observed in HRA between the preoperatively planned situation and the postoperatively reconstructed situation may be caused by the inclination and anteversion angles of the acetabular cups deviating from the ideal situation.The comparison of the three implant systems in the preoperatively planned situation (n = 19 patients per implant system) also showed physiological ROM 8,30,42,50,51 for the maximum flexion and internal rotation at 90°flexion movements.Analogous to the comparison between preoperatively planned and postoperatively reconstructed situation, ROM during external rotation at 90°flexion is overestimated for all three implant systems.In general, the ROM of the conventional hip stem and short hip stem during flexion was predominantly limited by bone-to-bone impingement, and only minor differences were found between the two implant systems.In contrast, the ROM of the short hip stem was increasingly limited by impingement with bone involvement (bone-to-bone and implant-to-bone) during internal rotation at 90°flexion and external rotation at 90°flexion.It is not only the implant-to-implant geometry but rather the positioning of the implant components in the bone structures that limits ROM at flexion. 24,60It may be possible that leaving a portion of the femoral neck in place due to the osteotomy of the short hip stem leads to increased bone impingement during internal and external rotation compared with the conventional hip stem.2][63][64] However, the risk of dislocation may increase with frequent bone-to-bone impingement, even with carefully placed implant components. 42The implant-to-bone impingement that occurred predominantly with hip resurfacing during all three performed movements can be attributed to the preservation of the femoral neck. 8The ROM achieved depends on the femoral neck configuration, which can result in unexpectedly reduced ROM despite large heads in combination with short neck lengths. 657][68] Regarding the ROM, HRA is correlated with reduced levels compared with conventional hip stems in previous studies. 3,8,26,69so in line with the current literature, 24 increased ROM for maximum flexion and internal rotation at 90°flexion were found with increasing head-to-neck ratio in the conventional hip stem and short hip stem.It was observed that the conventional hip stem tends to allow higher flexion and internal rotation for the same head sizes.In general, larger head diameters result in higher ROM, 23,24,44,47,55,65,70,71 which was observed especially for the conventional hip stem and the short hip stem endoprostheses.
Higher head offsets are known to increase the ROM. 19,24,27,65This behavior becomes apparent for the conventional hip stem and the short hip stem only from head size >32 mm.Regarding HRA, it should be mentioned that due to patient-specific differences, femoral neck geometry differs in length and diameter in all 19 patients.Thus, the head-to-neck ratio not only depends on the variation in head size but is also influenced by the patient-specific femoral neck diameters.
In conclusion, the treatment of the osteoarthritic hip joint with the femoral component of THR is a challenge in younger active patients concerning the preservation of the proximal femur for later revision surgery and the required ROM.In our present study, a comparison of three different implant systems showed that all systems achieve physiological ROM during maximum flexion and internal rotation at 90°flexion.During internal rotation, impingement with bone involvement was more likely with preservation of the proximal femur.Hence, HRA showed significantly reduced ROM.
During external rotation at 90°flexion, the computationallycalculated ROM is overestimated for all three implant systems due to the lack of soft tissue structures considered.However, a comparison of the implant systems between the preoperatively planned situation and the postoperatively reconstructed situation shows the reliability of our modeling approach, which is based on the generation of bone geometries from noninvasive MRI data sets.| 2515 T A B L E A 2 (Continued)