Can FEM-based preoperative planning for complex tibial plateau fractures improve surgical skills? A cadaver study

Abstract

Background

Complex tibial plateau fractures continue to pose a significant challenge for surgeons. In recent years, the widespread use of CT imaging has led to new insights leading to novel classifications that facilitate 360° stabilization techniques. Visualization in 3D has improved both fracture reduction and surgical outcomes. This study investigated whether preoperative planning of complex tibial plateau fracture fixation via finite element modeling (FEM) could enhance the fixation performance achieved by experienced surgeons and potentially improve outcomes for less experienced surgeons.

Methods

In twelve left cadaveric fresh-frozen human knees with intact soft tissue reproducible Schatzker type IV fractures with lateral depression were created. The samples were paired on the basis of bone mineral density and then randomly allocated into two groups. Six senior surgeons with extensive experience in the operative treatment of tibial plateau fractures performed two procedures: one using standard preoperative planning and one using FEM-optimized fixation planning. All fractures were stabilized with a medial locking plate and supplemental single screws when needed. The operation time, radiation dose and implant usage were documented. Surgeon mental workload was measured by the NASA task load index. Finally, the samples were biomechanically tested over four quasistatic load ramps from 10 to 200 N, followed by a cyclic sinusoidal load with increasing load level until failure. Failure was defined as either ≥ 5° varus/valgus malalignment or a vertical impression of the condyles ≥ 3 mm. The initial stiffness and load to failure were assessed via a 3D motion tracking system. Statistical analysis was conducted using Student’s t-tests.

Results

No significant differences were observed in terms of operative time or intraoperative radiation exposure. However, the NASA-TLX mental demand test revealed a statistically significant advantage for the FEM-planned group (33 ± 12.4 vs. 49 ± 8.6 (p = 0.043)), indicating a reduced cognitive load. Additionally, the FEM group exhibited superior biomechanical performance, with a higher load to failure of 1050 ± 535 N vs. 442 ± 226 N (p = 0.041).

Conclusion

This biomechanical feasibility study demonstrated that FEM-based preoperative planning is feasible and easy to implement for complex tibial plateau fractures. This planning supports specialized surgeons in challenging operations and can improve the stability of osteosynthesis.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13018-026-06728-0.

Background

Open reduction and internal fixation of intra-articular tibial plateau fractures is a cornerstone of orthopedic trauma care, aiming to restore anatomy, promote healing, and restore patient function by ensuring mechanical stability at the fracture site[1]. The complexity of bone geometry, individual patient variability, broad fracture variation, and diverse biomechanical conditions involved present significant challenges during preoperative planning for fracture fixation[2]. Traditional planning relies on the clinical experience of the surgeon, which may not fully account for the multifactorial biomechanical environment of complex tibial plateau fractures[3].

Finite element modelling (FEM), a computational modelling technique widely adopted in engineering, has become a promising tool in orthopedic trauma surgery[4, 5]. The FEM enables the detailed simulation of bone and implant mechanics by dividing anatomical structures into discrete elements, allowing the prediction of stress distribution, strain, fracture gap motion, and implant stability under physiological loads[6]. This level of biomechanical insight is not achievable with standard imaging or intraoperative estimation alone. FEM-based analysis supports the understanding of fracture morphology, implant positioning, and, most importantly, patient-specific preoperative planning[4, 6].

By integrating computed tomography (CT) data of a tibial plateau fracture, surgeons can use the FEM to plan for fracture reduction, compare fixation strategies, assess primary and secondary stability, and predict failure risk before the fracture enters the operating room[4]. Recent studies have demonstrated the clinical benefits of this approach, showing that computer-assisted preoperative planning combined with FEM can reduce intraoperative blood loss, operative time, and fluoroscopy use compared with conventional planning methods[3].

The ideal scenario would be to provide every surgeon with access to FEM-based planning tools to determine the most effective strategy for open reduction and internal fixation. The initial question is whether FEM-based planning offers advantages in surgical practice and improves the stability of osteosynthesis compared with conventional approaches?

Methods

Specimens

Twelve fresh-frozen cadaveric human knees, extending from the femoral shaft to the distal tibia with intact soft tissue, were used. The specimens were obtained from Science Care (Science Care, Phoenix, AZ, US) via Rimasys (Rimasys GmbH, Cologne, Germany). Donor consent for research purposes was obtained. The specimens were specifically prepared by Rimasys for this study as follows: a CT scan of the intact bone was performed to exclude specimens with osseous pathologies and to determine bone mineral density (BMD). Tibia trabecular bone mineral density was measured in the unfractured metaphyseal region just below the lateral plateau[7–9]. To account for the trabecular bone mineral density across the whole tibia, voxel-based histograms of Hounsfield Units were created where the median trabecular bone mineral density was also evaluated[10]. Using a standardized, patented machine-controlled axial loading protocol[11], Schatzker type IV [6]/AO Typ 41B3.3 [12]fractures with lateral depression were reproducibly induced in all the samples while maintaining an intact soft tissue mantle. From a selection of 18 induced fractures on left human cadaveric tibia the 12 most similar fractures were selected by two orthopaedic trauma surgeon.

On the basis of fracture patterns and bone mineral density, the samples were paired and then randomly allocated into the standard group or the FEM planned group (Table 1).

Table 1.

This table presents the paired fractures together with the corresponding bone density

Pair/Group	Conventional	Interventional
Pair 6 Case 6 Case 12
Lateral metaphysis	182 mgHA/cm3	110 mgHA/cm3
Median total tibia	143 mgHA/cm3	133 mgHA/cm3
Pair 5 Case 5 Case 11
Lateral metaphysis	147 mgHA/cm3	146 mgHA/cm3
Median total tibia	134 mgHA/cm3	151 mgHA/cm3
Pair 4 Case 2 Case 8
Lateral metaphysis	117 mgHA/cm3	180 mgHA/cm3
Median total tibia	164 mgHA/cm3	154 mgHA/cm3
Pair 3 Case 4 Case 10
Lateral metaphysis	59 mgHA/cm3	87 mgHA/cm3
Median total tibia	93 mgHA/cm3	104 mgHA/cm3
Pair 2 Case 3 Case 9
Lateral metaphysis	119 mgHA/cm3	63 mgHA/cm3
Median total tibia	131 mgHA/cm3	95 mgHA/cm3
Pair 1 Case 1 Case 7
Lateral metaphysis	121 mgHA/cm3	111 mgHA/cm3
Median total tibia	144 mgHA/cm3	148 mgHA/cm3
Overall lateral metaphysis (p = 0.727)	124 ± 37 mgHA/cm3	116 ± 38 mgHA/cm3
Overall median of the total tibia (p = 0.769)	135 ± 21 mgHA/cm3	131 ± 23 mgHA/cm3

The visualization of the fractures depicts the white and grey areas as the unaffected regions of the tibial plateau. The view shown is the axial perspective of the tibial plateau

Six senior trauma surgeons with extensive experience in the operative treatment of tibial plateau fractures (> 100 cases) were invited to perform two procedures: one with standard preoperative planning and one with FEM-based planning. The operations were conducted separately to avoid mutual influence, with a six month interval between procedures to prevent habituation to the test setup. For biomechanical comparability, all samples were stabilized using an anatomically pre-contoured locking plate for the medial condyle. The surgical approaches for fracture reduction were not specified. In the standard treatment group, the use of individual screws and their quantity were left to the surgeon’s discretion. The operations were performed under fluoroscopic control.

Implants

A polyaxial locking plate for the medial plateau was used (Pangea, proximal medial tibia plate, left, 8-hole, Stryker GmbH, Selzach, Switzerland). Implantation was performed according to the manufacturer's instructions. For additional fixation, surgeons could use 4.0 mm cannulated screws (ASNIS, Stryker GmbH, Selzach, Switzerland).

General procedure

The entire procedure was video recorded for later verification. The operation was divided into two steps: Step 1, from the skin incision to the beginning of osteosynthesis (reposition time), and Step 2, from the start of osteosynthesis to the final fluoroscopic images after placement of the last screw (osteosynthesis time). Skin closure was not performed. Surgeons could use a fluoroscope (Siremobil Compact L, Siemens Healthineers AG, Forchheim, Germany) whenever deemed necessary. The radiation dose was measured separately for each of the two steps.

Following the operation, surgeons completed a questionnaire that assessed mental workload via the NASA Task Load Index (NASA TLX)[13]. The NASA TLX is a subjective workload assessment tool developed by NASA`s Ames Research Center and is widely used in fields such as aviation, healthcare, military operations, and human–computer interaction to measure perceived mental, physical, and temporal demands, performance, effort, and frustration during a task.

Procedure 1: standard planned group

Surgeons were provided with radiological images of the fracture 24–48 h before the operation for review and surgical planning. Each fracture had native X-rays in two planes and a CT with axial, sagittal, coronal, and 3D reconstructions, with a slice thickness of 0.65 mm. On the day of surgery, surgeons can familiarize themselves with the implants and their application.

Procedure 2: FEM planned group

Each sample in Procedure 2 underwent additional surgical planning via Finite Element Modeling (FEM). Bone quality and fracture patterns were used to optimize the placement and orientation of the osteosynthesis plate and screws. Screw and plate positioning were optimized for maximal stability of the reconstruction[14] (Fig. 1). We minimized overall global displacement. Hounsfield-Unit-derived bone material properties as well as joint and muscle forces from subject-specific musculoskeletal gait models were integrated in the FEM [14].

Fig. 1.

Plate and screw positioning was optimized with the help of the FEM to achieve the highest possible stability (minimal displacement magnitude under loading in the FEM)

Surgeons were provided with CT scans of the fracture and the CustomSurg Ortho Planner for 3D visualization of the fracture, reduction, and preoperative planing with plate and screw positions (CustomSurg AG, Zurich, Switzerland) (Fig. 2). 3D-printed patient-specific models were also generated. For optimal plate and screw positions, customized 3D-printed targeting devices were supplied. (Fig. 3).

Fig. 2.

Rendering of the FEM planed screw and plate positioning after repositioning of the fragments

Fig. 3.

Specimen with the implanted plate and 3D-printed targeting devices

A deviation in screw placement was observed when the screw length used by the operating surgeon did not match the length planned in the FEM model; instead, a longer or shorter screw size was selected.

Biomechanical testing

After surgery, the soft tissues were removed and each tibia was cut to a length of 17 cm and embedded, ensuring horizontal alignment of the tibia plateau. An axial load was applied via artificial femoral condyles to the tibial plateau using an electrodynamic material testing machine (Instron E3000, Instron GmbH, Germany) (Fig. 4). Marker flags were attached to the fracture fragments for 3D motion tracking.

Fig. 4.

Test setup with an embedded tibia sample and load application via two femoral hemicondyles mounted on a rocker to avoid constraint forces

The load protocol was divided into three steps. First, the construct was pre-conditioned with sinusoidal axial load from 10 to 50 N at 1 Hz for 100 cycles. This was followed by four quasi-static displacement-controlled loading ramps at a velocity of 2 mm/min from 10 to 200 N, where the fourth ramp was used to measure axial construct stiffness. In a third step, cyclic sinusoidal loading started from 10 to 50 N at a frequency of 1 Hz. Every 500 cycles, the maximum load was incrementally increased by 50 N, while the lower load level was kept constant at 10 N. Failure was defined as either a vertical impression of the condyles exceeding 3 mm into the tibia plateau or a varus/valgus malalignment greater than 5° [15].

Data analysis

Initial stiffness, tibia plateau widening and load to failure were analyzed via a 3D motion tracking system (ARAMIS 6M, Carl Zeiss GOM Metrology GmbH, Germany).

All inspected data were tested for a normal distribution via the Kolmogorov–Smirnov test.

Statistical analysis was conducted via Student’s t-tests, and the level of significance was set to 0.05 (SPSS Statistics, version 26, IBM, US).

Results

Bone mineral density (BMD) was comparable between the two groups (124 ± 37 mg/ccm for the standard planned group, 116 ± 38 mg/ccm for the FEM-planned group; p = 0.743) (Table 2). In both groups, the medial plate was fixed with an average of 10.6 screws. The standard planned group used a minimum of 8 screws and a maximum of 13 screws, whereas the FEM-planned group used a minimum of 10 screws and a maximum of 12 screws. The number of anteroposterior (AP) screws in the FEM-planned group was slightly greater (1.8 screws vs 1 screw on average).

Table 2.

Inspected parameters for the standard planned group and the FEM-planned group with the respective p-values

	Standard planned	FEM planned	p-value
BMD mg/ccm2	124 ± 37	116 ± 38	0.743
Screws in total	10.2 (min 8 max 13)	10.7 (min 10 max 12)	1
AP Screws	1 (min 0 max 3)	1.8 (min 1 max 3)	0.141
Repostion in min	27 ± 13	18 ± 9	0.245
Osteosynthesis in min	29 ± 11	41 ± 11	0.112
Surgical time in min	56 ± 21	60 ± 8	0.732
Dose cGy reposition	148 ± 87	91 ± 55	0.246
Dose cGy osteosynthesis	339 ± 188	435 ± 77	0.313
NASA-TLX	49 ± 18.6	33 ± 12.4	0.043*
Stiffness in N/mm	650 ± 392	899 ± 445	0.37
Load to failure in N	442 ± 226	1050 ± 535	0.041*

The asterisk represents a significant difference

In the FEM-planned group, a total of 64 screws were planned in terms of direction and length, and 59 screws (92.2%) were placed in the designated positions with the predicted length. Consequently, 7.8% of the screws deviated. The total operation time was similar for the two groups (56 ± 21 min for the standard planned group vs. 60 ± 8 min for the FEM-planned group). The time for repositioning was slightly shorter in the FEM-planned group (18 ± 9 min) than in the standard planned group (27 ± 13 min; p = 0.245), although this difference was not statistically significant. Similarly, the radiation dose during repositioning was greater in the standard planned group (148 ± 87 cGy vs 91 ± 55 CGy; p = 0.246) and lower during osteosynthesis than in the FEM-planned group (339 ± 188 cGy vs 435 ± 77 cGy; p = 0.313).

In the postoperative questionnaire, the NASA- TLX revealed that the FEM-planned group (33 ± 12.4) experienced a significantly lower workload than did the standard planned group (49 ± 18.6) (p = 0.043).

Biomechanical testing revealed that the initial stiffness was greater in the FEM-planned group, although the difference was not statistically significant (899 ± 445 N/mm vs. 650 ± 392 N/mm; p = 0.370). Failure occurred at 442 ± 226 N for the standard planned group, whereas the FEM-planned group failed at significantly higher loads of 1050 ± 535 N (p = 0.041).

Discussion

In our experimental study, we demonstrated that FEM-assisted surgical planning combined with 3D printing technology and the use of customized targeting devices is feasible for managing complex Schatzker type IV/AO typ 41B3.3 tibial plateau fractures. To our knowledge, this is the first report utilizing FEM-based analysis as the foundation for individual osteosynthesis planning in the stabilization of realistic tibial plateau fractures in a human cadaveric model.

Three-dimensional (3D) printing technology has shown promising potential in the surgical management of tibial plateau fractures. Several studies have indicated that 3D-assisted surgery can reduce operation time, lower intraoperative blood loss, and decrease fluoroscopy exposure compared with conventional techniques[16] 17]. In addition, 3D printing applications may have the potential to accelerate fracture healing and lower complication rates[2]. Various 3D-assisted surgical methods have been developed, such as virtual visualizations, printed fracture models, precontoured plates, surgical guides, and intraoperative 3D imaging[18], 19, 20]. While some evidence suggests improved postoperative knee function with 3D-assisted surgery[2], other studies have reported no significant differences in functional outcomes or complication rates compared with conventional approaches[16, 17]. Overall, 3D printing technology appears to offer notable advantages in the surgical treatment of tibial plateau fractures, although adequate clinical trials are needed to confirm its superiority.

Our findings indicate that preoperative planning of fracture treatment via 3D printing technology can accelerate fracture reduction and potentially reduce radiation exposure.

However, the use of human bone in FEM planning remains challenging, particularly in fracture situations[14]. While biomechanical studies have confirmed the validity of FEM simulation combined with mechanical testing in models of proximal tibial fractures[21], most have used simplified osteotomies or synthetic bones. These fracture models provide reproducible and resource-efficient testing setups but limit their clinical transferability[22]. In contrast, the use of human cadaver models with in situ-created, reproducible Schatzker type IV fractures represents a significant step toward clinical reality.

A recognized limitation of biomechanical studies using human specimens is inter-specimen variability, particularly when fractures are created in situ. This variability may result in imbalances between experimental groups[23] making careful specimen allocation essential to reduce systematic errors. The biomechanical setup and finite element method (FEM) planning generate a substantial volume of additional data, the analysis of which will be investigated in subsequent studies.

We selected the Schatzker type IV or AO type 41B3.3 fracture pattern to create a highly unstable fracture involving both the medial tibial plateau and the lateral impression zone. Isolated medial fixation provided sufficient mechanical stability, for meaningful group comparability. Although other fracture patterns, such as Schatzker type II (AO type B3.1) and Schatzker types V and VI are more common [24], they are either more stable or show greater variability, requiring different surgical techniques. Thus, the Schatzker type IV fracture pattern was most suitable for our study.

Notably, that in some samples, Schatzker type IV fractures extend laterally, approaching bicondylar fracture patterns due to a broad impression zone. In clinical practice, such cases might prompt surgeons to consider additional lateral plating for safety[25].

In the FEM-planned group, a customized targeting device enabled the precise placement of the plate and screws, with over 90% of screws matching the preoperative planning regarding direction and length. The 3D-printed guide was unfamiliar to the participating surgeons, as such aids have thus far been used only sporadically in our clinical routine. Recently, this method was described for the stabilization of medial tibial plateau fractures[19]. This unfamiliarity may have contributed to a slightly longer time to conduct osteosynthesis and increased radiation exposure. However, as experience with these technologies grows, the associated learning curve is expected to shorten, potentially leading to even greater time savings in real-world settings, especially given the advanced planning of screw positions.

The NASA-TLX, which was originally developed for aviation tasks, has also been employed in complex surgical procedures[13]. In our study, the FEM-planned group demonstrated significantly lower NASA-TLX scores, indicating a reduced mental workload during surgery and highlighting the benefits of comprehensive preoperative planning.

Biomechanical testing revealed slightly greater construct stiffness for the FEM-planned group, but the difference was not statistically significant. This result is understandable, as the primary plate was applied medially in both groups, accounting for most of the stiffness of the construct. The principal difference emerged in the number of ap screws, screw placement and orientation, which became apparent under cyclic loading[26]. This was reflected in the significantly greater failure load for the FEM-planned group, indicating enhanced biomechanical stability. Given this, failure load may be a more relevant predictor of clinical outcomes.

Limitations

This is a biomechanical setup that allows conclusions to be drawn for clinical applications. However, these findings need to be validated in clinical trials. In vitro studies have an inherent weakness in that an in vivo situation comprising muscle forces cannot be simulated. The use of human cadaver models with in situ-created fractures likely introduced variability in the samples. Furthermore, the limited sample size of n = 6 in each group, which is determined by the complexity of the fracture model, may have limited the ability to detect additional mechanical advantages. Finally, the availability of experienced senior surgeons for these complex fracture scenarios is inherently limited.

Conclusion

Preoperative FEM-based planning for complex proximal tibial fractures is feasible and offers a promising strategy for reducing the operative time, minimizing the mental workload, enhancing biomechanical stability, and potentially improving the clinical outcome.

Supplementary Information

Below is the link to the electronic supplementary material.

Authors’ contributions

All authors made substantial contributions to the study conception and design. Material preparation and data collection were performed by R.P., S.S., S.C. and L.S. Data were analyzed by R.P S.S.. and P.A. and discussed with S.C.,L.S., T.Z., A.K. and B.S. The first draft of the manuscript was written by R.P. and revised by S.S., S.C.,L.S., T.Z., A.K., B.S. and P.A.. All authors read and commented on previous versions of the manuscript and approved the final manuscript.

Funding

No funding was received for this study.

Data availability

Data is provided within the manuscript or supplementary information files.

Declarations

Ethics approval and consent to participate

Consultation with the institutional review board of BG Trauma Hospital Murnau, Germany, confirmed that, according to applicable national regulations, no ethics application or consent to participate was required for this study. All surgeons participated in the study voluntarily. The specimens were obtained from Science Care USA via Rimasys GmbH, Cologne, Germany. Donor consent for research purposes was obtained.

The study was conducted in accordance with the principles of the Declaration of Helsinki.

Consent for publication

All authors have read and approved the final version of the manuscript and consent to its publication. They agree to be accountable for all aspects of the work, including its accuracy and integrity.

Competing interests

Authors S.C., L.S., T.Z., and A.K. are employees of CustomSurg AG, which provided the finite element model used in this study. Additionally, B.S. is employed by Rimasys GmbH, the company that supplied the fracture model.