Assessment of morphological differences of the proximal tibia in healthy knees: analysis of the 3-dimensional mathematical model

Jolanta Tomczyk; Joanna Taczała; Jacek Sawicki; Marcin Domżalski

doi:10.21037/qims-20-1094

Original Article

Assessment of morphological differences of the proximal tibia in healthy knees: analysis of the 3-dimensional mathematical model

Jolanta Tomczyk^{1^}, Joanna Taczała^{2^}, Jacek Sawicki^{2^}, Marcin Domżalski^{3^}

¹Department of Radiology, Isotopic Diagnostic and Therapy, Veteran’s Memorial Hospital Medical University of Lodz, Lodz, Poland; ²Institute of Materials Science and Engineering, Lodz University of Technology, Lodz, Poland; ³Department of Orthopedics and Traumatology. Veteran’s Memorial Hospital Medical University of Lodz, Lodz, Poland

Contributions: (I) Conception and design: All authors; (II) Administrative support: All authors; (III) Provision of study materials or patients: J Tomczyk, J Taczała; (IV) Collection and assembly of data: J Tomczyk, J Taczała; (V) Data analysis and interpretation: All authors; (VI) Manuscript writing: All authors; (VII) Final approval of manuscript: All authors.

^{^}ORCID: Jolanta Tomczyk, 0000-0002-8942-0233; Joanna Taczała, 0000-0002-3201-9182; Jacek Sawicki, 0000-0001-9147-7338; Marcin Domżalski, 0000-0003-1915-0773.

Correspondence to: Jolanta Tomczyk. Department of Radiology, Isotopic Diagnostic and Therapy, Veteran’s Memorial Hospital Medical University of Lodz, Poland Zeromskiego 113 Street, 90-549 Lodz, Poland. Email: jol.tomczyk@gmail.com.

Background: High tibial osteotomy and many orthopedic surgical procedures around the knee joint requires precise preoperative planning. In-depth knowledge of the tibial plateau morphology is necessary to limit intraoperative complications like lateral hinge fracture. No studies were exploring the differences in proximal tibia surface geometry, in regards to gender and laterality, using a mathematical model. The aim of our study was to assess morphological differences in healthy knees using a three-dimensional mathematical model.

Methods: Eighty-seven computed tomography examinations collected from 52 patients were selected for the study. The inclusion criteria were: age between 20 and 40 years, knee joint without visible deformities, no history of significant trauma to index knee, no history of systemic and chronic disorders. The average age of the included patients was 32.5±8.9 years old. For the calculation and comparisons, 45 right knee joints (18 females and 27 males) and 42 left knee joints (17 females and 25 males) were used.

Results: The male tibial plateau was much larger than the female one, for the right (P=0.001) and left knees (P=0.001). Male knees showed much bigger variability in two-dimensional tibial plateau dimensions especially for the left knees (P=0.001), and there was also a marked difference in variability between sides in males. Three-dimensional variability was significant for medial condyles for both genders. Male knees had a statistically bigger (P=0.04) tibial plateau surface area for all measured condyles.

Conclusions: The proximal tibial plateau showed in the designed mathematical models high variability in the two-dimensional and three-dimensional analysis. The males' knees presented great variability between sides and condyles. This finding must be considered during preoperative planning.

Keywords: Tibia; three-dimensional model (3D model); anatomical variations

Submitted Sep 25, 2020. Accepted for publication Mar 26, 2021.

doi: 10.21037/qims-20-1094

Introduction

There has been a long-standing interest in the geometric bilateral symmetry of the bones forming the knee joint (1-3). This interest is partly due to the symmetry postulation that is frequently made in clinical assessments and research studies. A common clinical practice is to use the contralateral unaffected side as a template for planning corrective osteotomy (4,5).

Many templates designed for surgical planning of the procedures around the knee are using the contralateral limb as a reference, considering that the left and right extremities are not significantly different. Moreover, these templates assume that there are insignificant differences between males’ and females’ proximal tibia morphology.

Little is known, how chronic, degenerative diseases may alter the geometry of the articular surfaces of the knee joint. Osteoarthritis (OA) of the knee is often a cause of considerable deformation of the proximal tibia and previous studies have shown that this joint is more commonly affected by degenerative changes that any other in the human motor system (6).

Since the first Jackson report (7) in 1958 high tibial osteotomy (HTO) has been recognized for its value for the treatment of symptomatic arthritis of the knee, especially in a younger, more active patient (6). The success of this procedure is based on detailed preoperative planning of axial deformity of the lower extremity and congruency of the knee joint. HTO is predominantly scheduled for patients with unicompartmental medial OA, and varus knees. Since HTO changes the slope of the tibial plateau a better insight into the anatomical variations of the tibia plateau between genders and sides is mandatory for proper surgical planning. Lack of adequate, anatomical—based preparation for HTO may result in an intra-operative lateral hinge or intraarticular fracture and treatment failure (8).

The precise model of the proximal tibial epiphysis is required for more morphological and biomechanical studies.

The first aim of this study was to design a mathematical model for 3D reconstruction of average proximal tibia separately for males and females, as well as for right and left limbs. The second purpose was to find differences between created models concerning gender and dominant and non- dominant limbs.

Methods

Data collection

Computed tomography (CT) scans of patients who had undergone CT scanning between 2015 to 2019 due to unrelated orthopedic reasons to the index knee joint (i.e., vascular indications, injuries to the opposite leg) were collected from the university hospital database. Becuase our study did not disclose any private petient data, individual consent for this retrospective analysis was waived. The inclusion criteria were: age between 20 and 40 years, knee joint without visible deformations, no history of significant trauma to index knee, no history of systemic and chronic disorders. Each CT scan was evaluated by a trained medical doctor for signs of radiological OA in the knee joint (such as joint space narrowing, subchondral sclerosis, subchondral cyst formation and osteophytes). Any CT scan exhibiting signs of OA was excluded from the study. All inclusion criteria fulfilled 87 CT examinations collected from 52 patients out of 454 individuals screened for inclusion/exclusion criteria. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). An approval from the Local University Ethical Committee was obtained for the study design and data acquisition from CT scans. Because our study did not disclose any private patient data, individual consent for this retrospective analysis was waived.

The average age of the included patients was 32.5±8.9 years old. There were 33 men and 19 women in the study group. For the calculation and comparisons, 18 females and 27 males right knee joints, and 17 females and 25 males left knee joints, were used.

Imaging

The CT scans were made using a Siemens SOMATOM Sensation 64 CT scanner. The acquisition parameters were the following: tube voltage 120 kVp, with automatic selection of exposure conditions-CARE DOSE, slice thickness 1 mm, a matrix size of 512×512 pixels, and a pixel size of 1.52344 mm, rotation time 1 s, Pitch 0,55.

The first part of the work was transforming 2D CT images into 3D geometry by the image segmentation process. This action was intended to obtain a virtual 3D model named later as CAD body (CAD-computer-aided design). Later, the models were edited in 3D modeling software to obtain a closed model needed for numerical calculations.

Then with GOM Inspect V8 SR1 (Braunschweig, Germany) and ANSYS 2019 R3 software (by ANSYS Inc. company), 3D models were analyzed. Right and left tibial bones as well as the medial and lateral tibial condyle plateau were analyzed separately.

Bone width

First, the bone width was measured in the longest line in the cross-section view at the midplane at the coronal view (Figure 1). The width was calculated separately for each bone using GOM Inspect software.

Figure 1 Exemplary place of measurement.

2D dimensions analysis

3D models were compared to the CAD body of most average bone (selected using geometry pre-adjustment). Then cross-section plane was created in the measured center of base bone in the Y-axis view. Differences in the shape in Z-axis were measured by comparing other bones to the base one at the line of the created cross-section.

Models alignment was automatically calculated by the software, according to the middle line. For every measurement, the result was a set of points at equal distance (0.50 mm) in the X-axis (Figure 2). All measured points were in the same position according to gender and side of the body (right/left leg).

Figure 2 Example of a set of measurements points in 2D cross-section.

A colorful map of deviations was created for better visualization (legend on the right side in Figure 2).

3D dimensions analysis

The differences in dimension over the entire surface of the plateau were measured by generating a deformation map on whole geometry. Again, in this case, all 3D models were compared to the basic one. Then the set of measurement points was generated and calculated separately for the medial and lateral plateau. In this case measurement points were created at an equal distance of 1.20 mm (Figure 3).

Figure 3 Example of a map of deformation with the cloud of measurements points.

Plateau surface area

The surface area for each plateau separately was measured using the SpaceClaim module for ANSYS software. The surface area was determined according to the highest points of geometry as compared to the lowest points in the plateau area (Figure 4).

Figure 4 Example of measurement of plateau surface area.

The angle of inclination of the plateau

To measure the angle of inclination of the plateau at first bones were aligned like in section “2D dimensions analysis” and “3D dimensions analysis”. The orientation of the reference plane is based on the center axis of the bone. Two reference planes were added—one in ZY-axis (Figure 5A) and second—in ZX-axis (Figure 5B,C). The last step was the measurement of the angle between the reference plane and planes of the plateau. An example of the measured angles is shown in Figure 6. The procedure was the same for every single 3D mesh. Every comparison used the same reference plane.

Figure 5 Example of measurement of inclination of the plateau: (A) sides, (B) front, (C) back.

Figure 6 Example of measurement of inclination of plateau.

Statistical analysis

Data obtained by measurements were tested using a single factor ANOVA test statistical significance assumed at a level of P<0.05. Statistical analysis was the determination of mean and median and standard deviation (SD). The fille area limits are the minimum and maximum values.