Mediolateral femoral component position in TKA significantly alters patella shift and femoral roll-back

Is Mediolateral Translation of Femur Caused by Total Knee Arthroplasty Being Ignored?

Purpose

This study aims to show the change in overall congruency due to mediolateral translation after total knee arthroplasty compared with normal knee anatomy.

Methods

This study was performed in two parts. In part 1, the relationship between femur and tibia was defined by new parameters on the antero-posterior radiographs of 84 patients. In part 2, this relationship was evaluated on the postoperative radiographs of 136 total knee arthroplasty patients. Two parallel lines to the tibial anatomical axis were drawn tangent to the most lateral and most medial parts of the tibial plateau. After creating medial and lateral tangential lines, the distance between the most lateral point of the lateral femoral epicondyle and lateral tangential line and the most medial point of the medial femoral epicondyle and medial tangential line was measured. Another new parameter described in the study is epicondylar distance ratio. The ratios between the shortest distance between tibial anatomical axis and lateral femoral epicondyle and the distance between tibial anatomical axis and medial femoral epicondyle were defined.

Results

It was found that the lateral tangent was not superposed in any measurement to the femoral lateral condyle, the closest tangent was passed, and the mean lateral space distance was 1.8 mm (SD 1.5, 95% CI 0-5.3 mm). The medial tangent was passed from the lateral to the femoral medial epicondyle, and the medial crossing distance was 8.5 mm (SD 5.7, 95% CI 5-14 mm). Epicondylar distance ratio used as the second measurement was 0.8 (0.5-0.9). After total knee arthroplasty measurements showed that the line passing through the lateral tibia crossed the lateral epicondyle of the femur and intersected at an average distance of 4.3 mm (SD 4.1, 95% CI 1-11.2 mm).

Conclusions

There is a coronal plane congruence between tibia and femur in the healthy knees, which get changed after total knee arthroplasty.

Effects of Femoral Component Design on the Deepest Point Position of the Trochlear Grove in Kinematically Aligned Total Knee Arthroplasty: A Comparison of Four Prothesis Designs

The aim of the study is to explore and compare the differences in trochlear shape and knee anatomy between four types of prostheses and preoperative native knee matched with preoperative computed tomography (CT). Thirty patients were scheduled for primary kinematically aligned total knee arthroplasty (TKA) for varus knee osteoarthritis at our hospital and the region between their pelvis to ankle joint was simulated using a CT-based three-dimensional planning software. The axial plane containing the transepicondylar axis was set as Slice A, and the 10-mm distal plane from Slice A was set as Slice B. The distances to the deepest trochlear groove between the native knee and each prosthesis and the medial and lateral facet heights were compared among the four groups. The deepest femoral trochlear groove of the prostheses was located 1.6 to 3.0 mm more medial than that of the native knee, and in the Persona group, it was significantly more medial than in the e-motion or Triathlon groups on both Slices A and B. The native knee and the medial and lateral facet heights of the four prostheses on both Slices A and B were significantly lower than those of preoperative native knees when femoral prostheses were set in the kinematically aligned (KA)-TKA position. The deepest point of the trochlear groove of the Persona group was the most medial among the four prostheses studied, and the deepest points differed depending on the prosthesis design in KA-TKA. Thus, surgeons should carefully select the type of prostheses used in KA-TKA.

Influence of kinematic alignment on femorotibial kinematics in medial stabilized TKA design compared to mechanical alignment

INTRODUCTION

Worldwide more and more primary knee replacements are being performed. Kinematic alignment (KA) as one of many methods of surgical alignment has been shown to have a significant impact on kinematics and function. The aim of the present study was to compare KA and mechanical alignment (MA) with regard to femorotibial kinematics.

MATERIALS AND METHODS

Eight fresh frozen human specimens were tested on a knee rig during active knee flexion from 30 to 130°. Within the same specimen a medial stabilized (MS) implant design was used first with KA and then with MA.

RESULTS

The femorotibial kinematics showed more internal rotation of the tibia in KA compared to MA. At the same time, there was a larger medial rotation point in KA. Both alignment methods showed femoral rollback over the knee bend.

CONCLUSION

Relating to an increased internal rotation and a more precise medial pivot point, it can be concluded that KA combined with a MS implant design may partially support the reproduction of physiological knee joint mechanics.

Computer-based analysis of different component positions and insert thicknesses on tibio-femoral and patello-femoral joint dynamics after cruciate-retaining total knee replacement

BACKGROUND

Positioning of the implant components and tibial insert thickness constitute critical aspects of total knee replacement (TKR) that influence the postoperative knee joint dynamics. This study aimed to investigate the impact of implant component positioning (anterior-posterior and medio-lateral shift) and varying tibial insert thickness on the tibio-femoral (TF) and patello-femoral (PF) joint kinematics and contact forces after cruciate-retaining (CR)-TKR.

METHOD

A validated musculoskeletal multibody simulation (MMBS) model with a fixed-bearing CR-TKR during a squat motion up to 90° knee flexion was deployed to calculate PF and TF joint dynamics for varied implant component positions and tibial insert thicknesses. Evaluation was performed consecutively by comparing the respective knee joint parameters (e.g. contact force, quadriceps muscle force, joint kinematics) to a reference implant position.

RESULTS

The PF contact forces were mostly affected by the anterior-posterior as well as medio-lateral positioning of the femoral component (by 3 mm anterior up to 31 % and by 6 mm lateral up to 14 %). TF contact forces were considerably altered by tibial insert thickness (24 % in case of + 4 mm increase) and by the anterior-posterior position of the femoral component (by 3 mm posterior up to 16 %). Concerning PF kinematics, a medialised femoral component by 6 mm increased the lateral patellar tilt by more than 5°.

CONCLUSIONS

Our results indicate that regarding PF kinematics and contact forces the positioning of the femoral component was more critical than the tibial component. The positioning of the femoral component in anterior-posterior direction on and PF contact force was evident. Orthopaedic surgeons should strictly monitor the anterior-posterior as well as the medio-lateral position of the femoral component and the insert thickness.

Prospective sequential comparison of femoral roll-back between cruciate-retaining and posterior-stabilized total knee arthroplasty using an intra-operative sensor

BACKGROUND

Implant design and surgical techniques affect postoperative knee kinematics in total knee arthroplasty (TKA). This study aimed to compare femoral roll-back between cruciate-retaining (CR) and posterior-stabilized (PS) TKA in the same knee by objectively quantifying the contact point kinematics of the tibiofemoral joint using a sensor.

METHODS

In the present prospective study, we used an intraoperative sensor to compare medial and lateral roll-back during 0-120° knee flexion in 33 knees that underwent CR and PS TKA. We also examined the relationship between mediolateral balance and the lateral-to-medial roll-back ratio. We defined the contact percentage position as the vertical length to the contact point divided by the anteroposterior length of the tibial plate.

RESULTS

The roll-back percentage following PS TKA (19.8 ± 5.1%) was significantly higher than that after CR TKA in both the medial (19.8 ± 5.1% versus 7.1 ± 2.5%, P < 0.001) and lateral (26.8% ± 3.8% versus 18.7 ± 3.8%, P < 0.001) compartments. The medial contact pressure at 90° was significantly correlated with the increased lateral-to-medial roll-back ratio in both CR and PS TKA (both P < 0.001).

CONCLUSION

PS TKA resulted in a higher percentage of femoral roll-back in the medial and lateral compartments than CR TKA. CR TKA caused a higher lateral-to-medial roll-back ratio compared to PS TKA. To reproduce medial pivot knee motion similar to that of a normal knee, the medial soft tissue needed to be balanced more tightly than the lateral soft tissue during TKA. These findings provide some clinical evidence of TKA design selection and proper mediolateral balancing for successful TKA.

Does Posterior Tibial Slope Influence Knee Kinematics in Medial Stabilized TKA?

BACKGROUND

During total knee arthroplasty (TKA), one of the key alignment factors to pay attention to is the posterior tibial slope (PTS). The PTS clearly influences the kinematics of the knee joint but must be adapted to the coupling degree of the specific TKA design. So far, there is hardly any literature including clear recommendations for how surgeons should choose the PTS in a medial stabilized (MS) TKA. The aim of the present study is to investigate the effects of different degrees of PTS on femorotibial kinematics in MS TKA.

MATERIALS AND METHODS

An MS TKA was performed in seven fresh-frozen human specimens successively with 0°, 3°, and 6° of PTS. After each modification, weight-bearing deep knee flexion (30-130°) was performed, and femorotibial kinematics were analyzed.

RESULTS

A lateral femoral rollback was observed for all three PTS modifications. With an increasing PTS, the tibia was shifted more anteriorly on the lateral side (0° PTS anterior tibial translation -9.09 (±9.19) mm, 3° PTS anterior tibial translation -11.03 (±6.72) mm, 6° PTS anterior tibial translation 11.86 (±9.35) mm). No difference in the tibial rotation was found for the different PTS variants. All PTS variants resulted in internal rotation of the tibia during flexion. With a 3° PTS, the design-specific medial rotation point was achieved more accurately.

CONCLUSIONS

According to our findings, we recommend a PTS of 3° when implanting the MS prosthesis used in this study.

Lateral shift of the femoral condyle after total knee arthroplasty: simulation using 2D-templates of the medial pivot design on knee radiographs of young Japanese patients

Background

Total knee arthroplasty (TKA) is an established surgical treatment for advanced knee osteoarthritis by which patients can expect improvement of knee pain and function. Although many surgeons have investigated limb alignment after TKA, changes in coronal positional relation between the femur and tibia are not known well.

Methods

Radiographs of 105 knees of young Japanese patients between 20 and 49 years-old (60 men and 45 women) without osteoarthritic changes who received arthroscopic surgeries at our hospital were used in this study. Using 2D-templates of the medial pivot design (the FINE total knee), we simulated TKA on a SYNAPSE-PACS software. First, the femoral component was placed in normal knee alignment and then was merged to the medial concave of the insert where the tibial component was placed in neutral alignment. The length of the mediolateral shift of the femoral component was measured as an estimate of lateral shift of the femoral condyle, of which association with radiographic parameters including the femorotibial angle (FTA), lateral distal femoral angle (LDFA), and medial proximal tibial angle (MPTA) was analyzed. Subjects were classified into three groups according to the femoral component size that was chosen in simulation of TKA, and the lateral shift of the femoral condyle was compared between groups.

Results

The estimated mean lateral shift of the femoral condyle was 5.99 ± 1.98 mm and was greater in males than females (p < 0.05). Also, it was most highly correlated with the medial proximal tibial angle (MPTA) (r = − 0.553, p < 0.01). A group receiving larger component sizes significantly shifted more laterally compared with a group receiving smaller component sizes (p < 0.01).

Conclusions

These results suggest that the coronal positional relation between the femur and tibia is altered and subsequent ligament imbalance may occur after mechanically aligned TKA using the medial pivot design.

Impact of femoro-tibial size combinations and TKA design on kinematics

INTRODUCTION

The variability in patients' femoral and tibial anatomy requires to use different tibia component sizes with the same femoral component size. These size combinations are allowed by manufacturers, but the clinical impact remains unclear. Therefore, the goals of our study were to investigate whether combining different sizes has an impact on the kinematics for two well-established knee systems and to compare these systems' kinematics to the native kinematics.

MATERIALS AND METHODS

Six fresh frozen knee specimens were tested in a force controlled knee rig before and after implantation of a cruciate retaining (CR) and a posterior-stabilized (PS) implant. Femoro-tibial kinematics were recorded using a ultrasonic-based motion analysis system while performing a loaded squat from 30° to 130°. In each knee, the original best fit inlay was then replaced by different inlays simulating a smaller or bigger tibia component. The kinematics obtained with the simulated sizes were compared to the original inlay kinematics using descriptive statistics.

RESULTS

For all size combinations, the difference to the original kinematics reached an average of 1.3 ± 3.3 mm in translation and - 0.1 ± 1.2° in rotation with the CR implant. With the PS implant, the average differences reached 0.4 ± 2.7 mm and  - 0.2 ± 0.8°. Among all knees, no size combination consistently resulted in significantly different kinematics. Each knee showed a singular kinematic pattern. For both knee systems, the rotation was smaller than in the native knee, but the direction of the rotation was preserved. The PS showed more rollback and the CR less rollback than the native knee.

CONCLUSION

TKA systems designed with a constant tibio-femoral congruency among size combinations should enable to combine different sizes without having substantial impact on the kinematics. The rotational pattern was preserved by both TKA systems, while the rollback could only be maintained by the PS design.

Secondary Patellar Resurfacing in TKA: A Combined Analysis of Registry Data and Biomechanical Testing

The German Arthroplasty registry (EPRD) has shown that different prosthesis systems have different rates of secondary patellar resurfacing: four years after implantation, the posterior-stabilized (PS) Vega prosthesis has a 3.2% risk of secondary patellar resurfacing compared to the cruciate-retaining (CR) Columbus prosthesis at 1.0% (both Aesculap AG, Tuttlingen, Germany). We hypothesized that PS implants have increased retropatellar pressure and a decreased retropatellar contact area compared to a CR design, which may lead to an increased likelihood of secondary patellar resurfacing. Eight fresh frozen specimens (cohort 1) were tested with an established knee rig. In addition, a possible influence of the registry-based patient collective (cohort 2) was investigated. No significant differences were found in patient data–cohort 2-(sex, age). A generally lower number of PS system cases is noteworthy. No significant increased patella pressure could be detected with the PS design, but a lower contact area was observed (cohort 1). Lower quadriceps force (100°–130° flexion), increased anterior movement of the tibia (rollback), greater external tilt of the patella, and increasing facet pressure in the Vega PS design indicate a multifactorial cause for a higher rate of secondary resurfacing which was found in the EPRD patient cohort and might be related to the PS’ principle function.

A Surgical Technique to Avoid Mediolateral Overhang during Total Knee Arthroplasty

Femoral component sizing and positioning is an essential component of performing a successful total knee arthroplasty. Failure to size or position the femoral component correctly can result in reduced range of motion, instability, soft-tissue impingement, and irritation. The following is a description of a novel technique that allows the surgeon to use an intercondylar autograft to translate the femoral component in the mediolateral direction and avoid the consequences of mediolateral femoral component overhang.

The tibial cut influences the patellofemoral knee kinematics and pressure distribution in total knee arthroplasty with constitutional varus alignment

PURPOSE

The current literature suggests that kinematic total knee arthroplasty (kTKA) may be associated with better outcome scores in patients with constitutional varus alignment. The underlying patellofemoral kinematic changes (patella tilting and patella tracking) and patellofemoral pressure distribution have not yet been described. The present study compared the effects of different tibial cuts, as used in kTKA, on patellofemoral knee kinematics and the pressure distribution, in addition to comparisons with the natural constitutional varus knee.

METHODS

Seven cadaveric knee joints with constitutional varus alignment were examined in the native state and after 0°, 3°, or 6° tibial cut cruciate-retaining (CR)-TKA using an established knee joint simulator. The effects on patella rotation/patella tilting, patellofemoral pressure, and patellofemoral length ratios (= patella tracking) were determined. In addition, the natural knee joint and different tibial cuts in CR-TKA were compared (Student's t test).

RESULTS

In the patellofemoral joint, 6° CR-TKA was associated with the greatest similarity with the natural constitutional varus knee. By contrast, knees subjected to 0° CR-TKA exhibited the largest deviations of patellar kinematics. The smallest difference compared with the natural knee joint concerning patella tilting was found for 6° CR-TKA (mean 0.4°, p < 0.001), and the largest difference was noted for 0° CR-TKA (mean 1.7°, p < 0.001). Concerning patellofemoral pressure, 6° CR-TKA resulted in outcomes most similar to the natural knee joint, featuring a mean difference of 3 MPa. The largest difference from the natural knee joint was identified for 0° CR-TKA, with an average difference of 8.1 MPa (p < 0.001; total mean 17.7 MPa). Meanwhile, 3° and 6° CR-TKA induced medialization of the patella, with the latter inducing the largest medialization value of 4.5 mm at 90° flexion.

CONCLUSIONS

The improved outcome parameters in kTKA described in the literature could be attributable to the similar kinematics of the patellofemoral joint relative to the normal state. The current study confirmed the similar kinematics between the native constitutional varus knee joint and knee joints subjected to 3° or 6° CR-TKA (patellofemoral rotation/patella tilting and patella pressure). Conversely, there was pronounced medialization of the patella following 6° CR-TKA. Patella pressure and patella tilting are described in the literature as possible causes of anterior knee pain after TKA, whereas medialization of the patella, which is also influenced by other causes, might play a subordinate role.

LEVEL OF EVIDENCE

V, Biomechanical study.

Musculoskeletal Multibody Simulation Analysis on the Impact of Patellar Component Design and Positioning on Joint Dynamics after Unconstrained Total Knee Arthroplasty

Patellofemoral (PF) disorders are considered a major clinical complication after total knee replacement (TKR). Malpositioning and design of the patellar component impacts knee joint dynamics, implant fixation and wear propagation. However, only a limited number of studies have addressed the biomechanical impact of the patellar component on PF dynamics and their results have been discussed controversially. To address these issues, we implemented a musculoskeletal multibody simulation (MMBS) study for the systematical analysis of the patellar component’s thickness and positioning on PF contact forces and kinematics during dynamic squat motion with virtually implanted unconstrained cruciate-retaining (CR)-TKR. The patellar button thickness clearly increased the contact forces in the PF joint (up to 27%). Similarly, the PF contact forces were affected by superior–inferior positioning (up to 16%) and mediolateral positioning (up to 8%) of the patellar button. PF kinematics was mostly affected by the mediolateral positioning and the thickness of the patellar component. A medialization of 3 mm caused a lateral patellar shift by up to 2.7 mm and lateral patellar tilt by up to 1.6°. However, deviations in the rotational positioning of the patellar button had minor effects on PF dynamics. Aiming at an optimal intraoperative patellar component alignment, the orthopedic surgeon should pay close attention to the patellar component thickness in combination with its mediolateral and superior–inferior positioning on the retropatellar surface. Our generated MMBS model provides systematic and reproducible insight into the effects of patellar component positioning and design on PF dynamics and has the potential to serve as a preoperative analysis tool.

Impact of tibial baseplate malposition on kinematics, contact forces and ligament tensions in TKA: A numerical analysis

PURPOSE

Malposition of implant components in total knee arthroplasty (TKA) has consequences on tibiofemoral kinematics, contact forces and ligament tensions. To evaluate the impact of tibial baseplate malpositioning in the same knee, we conducted a computer simulation.

METHODS

An established weight-bearing finite element model of a fixed bearing TKA was used for the computer simulation. To evaluate the influence of tibial baseplate malposition, calculations were consecutively performed in neutral position, at 3° and 6° of internal and external rotation and at 3 mm and 6 mm of medial and lateral translation.

RESULTS

The highest effect of malposition was observed for ligament tensions, with a tendency of a greater influence for the 6 mm translation compared to 6° of rotation. Changes in contact forces and tibiofemoral kinematics were according to the alterations of ligament tensions. The highest ligament tension, contact force and femoral roll-back were registered for 6 mm medialization of the tibial baseplate.

DISCUSSION

Tibial baseplate malposition effects ligament tensions, tibiofemoral contact forces and kinematics and has a risk of unfavorable clinical results due to postoperative pain, reduced range of motion, instability and a higher rate of early loosening. Therefore, surgeons should aim for a neutral position of the tibial baseplate.

Kinematically aligned total knee arthroplasty reproduces more native rollback and laxity than mechanically aligned total knee arthroplasty: A matched pair cadaveric study

BACKGROUND

A growing body of evidence supports that kinematically aligned (KA) total knee arthroplasty (TKA) provides superior clinical outcomes and satisfaction than mechanically aligned (MA) TKA. In theory, KA TKA would restore knee kinematics closer to the native condition than MA TKA, but the current biomechanical evidence is lacking.

HYPOTHESIS

KA TKA would restore knee biomechanics to the native condition better than MA TKA.

METHODS

Seven pairs of cadavers were tested. For each pair, one knee was randomly assigned to KA TKA and the other to MA TKA. During KA TKA, the sizes of femur and tibia resections were equivalent to implant thickness to align with the patient-specific joint line. MA TKA was performed using conventional measured resection techniques. All specimens were mounted on a customized knee-testing system and digitized. Knee motions measured during flexion included rollback, axial tibiofemoral rotation, and laxities, specifically varus-valgus laxity, anterior-posterior translation, and internal-external rotation.

RESULTS

The pattern of knee motion following KA TKA was similar to the native knee. However, following MA TKA, both medial and lateral rollback and tibiofemoral axial rotation were decreased relative to those of the native knee. Valgus laxity was restored only after KA TKA, whereas varus laxity was restored only after MA TKA. Anterior translation was increased regardless of the alignment strategy. In addition, rotational laxities were restored after KA TKA, but external rotation laxity increased after MA TKA.

CONCLUSION

KA TKA restores femoral rollback and laxity to the native condition better than MA TKA. KA TKA may enhance functional performance and provide a more normal knee sensation.

LEVEL OF EVIDENCE

II, Controlled laboratory study.

Kinematically aligned total knee arthroplasty reproduces native patellofemoral biomechanics during deep knee flexion

PURPOSE

The implant positioning for kinematically aligned total knee arthroplasty (TKA) differs fundamentally from conventional mechanically aligned TKA. This difference may affect patellofemoral (PF) biomechanics after TKA. This cadaveric study tested the hypothesis that kinematically aligned TKA would restore PF biomechanics to the native condition better than mechanically aligned TKA.

METHODS

Seven pairs (14 knees) of fresh-frozen cadavers were tested. All specimens were mounted on a customized knee-testing system and digitized using a Microscribe 3DLX instrument (Revware Inc., Raleigh, NC, USA) to measure patellar kinematics in terms of patellar varus/valgus rotation, medial/lateral position, flexion/extension rotation and proximal/distal position at knee flexion angles of 0°, 30°, 60°, 90°, and 120°. The medial and lateral PF joint contact pressure distributions at 120° of knee flexion were measured using a K-scan system (Tekscan Inc., Boston, MA, USA). All patellae remained unresurfaced. For each pair, one knee was randomly assigned to kinematically aligned TKA and the other to mechanically aligned TKA performed using the conventional measured resection technique. During kinematically aligned TKA, the amount of femur and tibia resected was equivalent to implant thickness to maintain the patient-specific joint line. All patellar kinematics were measured and compared between the native condition and after surgery.

RESULTS

The patellae of mechanically aligned TKA rotated more valgus and was positioned more laterally compared with those of kinematically aligned TKA at knee flexion angles ≥ 90°. Neither the patellar flexion/extension rotation nor the proximal/distal position differed between either prosthetic knee design and the native knee at all flexion angles. The contact pressure distribution between the medial and lateral PF joint after kinematically aligned TKA were similar to those of the native knee, while the lateral PF joint contact pressure after mechanically aligned TKA was higher than that of the native knee (p = 0.038).

CONCLUSIONS

Kinematically aligned TKA better restores patellar kinematics and PF contact pressure distribution to the native condition than mechanically aligned TKA during deep knee flexion. These findings provide clues to understand why kinematically aligned TKA is associated with less anterior knee pain and better PF functional performance compared to mechanically aligned TKA. Patients undergoing kinematically aligned TKA may experience a more normal feeling during deep knee flexion activities.

Medial stabilized and posterior stabilized TKA affect patellofemoral kinematics and retropatellar pressure distribution differently

PURPOSE

Patellofemoral kinematics and retropatellar pressure distribution change after total knee arthroplasty (TKA). It was hypothesized that different TKA designs will show altered retropatellar pressure distribution patterns and different patellofemoral kinematics according to their design characteristics.

METHODS

Twelve fresh-frozen knee specimens were tested dynamically in a knee rig. Each specimen was measured native, after TKA with a posterior stabilized design (PS) and after TKA with a medial stabilized design (MS). Retropatellar pressure distribution was measured using a pressure sensitive foil which was subdivided into three areas (lateral and medial facet and patellar ridge). Patellofemoral kinematics were measured by an ultrasonic-based three-dimensional motion system (Zebris CMS20, Isny Germany).

RESULTS

Significant changes in patellofemoral kinematics and retropatellar pressure distribution were found in both TKA types when compared to the native situation. Mean retropatellar contact areas were significantly smaller after TKA (native: 241.1 ± 75.6 mm², MS: 197.7 ± 74.5 mm², PS: 181.2 ± 56.7 mm², native vs. MS p < 0.001; native vs. PS p < 0.001). The mean peak pressures were significantly higher after TKA. The increased peak pressures were however seen in different areas: medial and lateral facet in the PS-design (p < 0.001), ridge in the MS design (p < 0.001). Different patellofemoral kinematics were found in both TKA designs when compared to the native knee during flexion and extension with a more medial patella tracking.

CONCLUSION

Patellofemoral kinematics and retropatellar pressure change after TKA in different manner depending on the type of TKA used. Surgeons should be aware of influencing the risks of patellofermoral complications by the choice of the prosthesis design.