Improvement of arthroscopic surgical performance using a new wide-angle arthroscope in the surgical training.

BACKGROUND: We have developed a new arthroscope with a field of view of 150°. This arthroscope requires less motion to maneuver and exhibits reduced optical error. It also improves how novices learn arthroscopy. We hypothesized that the surgical performance with this arthroscope is superior to that with a conventional arthroscope. This study tested the hypothesis by using motion analysis and a new validated parameter, "dimensionless squared jerk" (DSJ). METHODS: We compared the surgical performance between the use of the wide-angle arthroscope and that of the conventional arthroscope among 14 novice orthopedic residents who performed 3 standardized tasks 3 times with each arthroscope. The tasks simulated the surgical skills in arthroscopic rotator cuff repair. The arthroscope motion was analyzed using an optical tracking system. The differences in performance parameters, such as the time taken to complete the tasks, average acceleration of the hands (m/s2), number of movements, and total path length (m) including DSJ between the 2 arthroscopes were investigated using paired t-tests. RESULTS: All estimated values for the tasks using the 150° arthroscope were lower than those for the tasks using the 105° arthroscope. Statistically significant differences in performance between the 2 arthroscopes were observed only for DSJ (p = 0.014) and average acceleration (p = 0.039). CONCLUSIONS: DSJ and average acceleration are reliable parameters for representing hand-eye coordination. The surgical performance of novice arthroscopists was better with the new wide-angle arthroscope than with the conventional arthroscope.

Comparison of the biomechanics of radial head prostheses with dynamic loading in the radiocapitellar joint.

We used eight fresh cadaveric elbows to evaluate the biomechanical characteristics of the native radial head, an anatomic radial head prosthesis, and a non-anatomic radial head prosthesis using a dynamic model. The biceps, brachialis, and triceps were attached to pneumatic actuators loaded to mimic muscle force. The radiocapitellar contact pressure and area were measured in real time, and the associated curves were depicted simultaneously. No significant differences in the contact area or associated curves were found between native radial head and the anatomic radial head prosthesis. The contact pressure and associated curve for the anatomic radial head prosthesis were better than those for the non-anatomic radial head prosthesis. We conclude from this study that the contact pressure and area of the elbow with anatomic radial head prosthesis are similar to those of the native radial head. The anatomic radial head prosthesis appears to be more suitable in clinical use.

Contact Mechanics of Anatomic Radial Head Prosthesis: Comparison Between Native Radial Head and Anatomic Radial Head Prostheses in the Dynamic Mode.

PURPOSE: The biomechanical characteristics of anatomic radial head prostheses have not been completely investigated. We compared and analyzed the contact kinematic characteristics of the native radial head and radial head prostheses replicating the native head contour, using a real-time flexion simulation model. METHODS: Ten fresh-frozen cadavers were used in this pilot study. A simulating dynamic motion mode from 0° to 130° of elbow flexion was applied. Radiocapitellar contact pressure and area were measured using a real-time digitized pressure sensor. Further, contact area and pressure curves were obtained during flexion, using a motion-tracking device. RESULTS: The mean contact area, mean contact pressure, and peak contact pressure of the native radial head and radial head prosthesis were 39 mm2, 0.0078 kgf/mm2, 0.0123 kgf/mm2, and 33 mm2, 0.0093 kgf/dm2, 0.0148 kgf/mm2, respectively. The contact area and pressure curves were parabolic nonlinear for the radial head prosthesis and more linear for the native radial head. CONCLUSIONS: The radial head prosthesis mimics the mechanics of the native radial head in terms of mean contact area, mean contact pressure, and peak contact pressure; however, different patterns of contact pressure and area curves during elbow flexion-extension were observed. CLINICAL RELEVANCE: We found that the biomechanics of the anatomic radial head prostheses used in the study were similar to those of the native radial head.

The Dimensionless Squared Jerk: An Objective Parameter That Improves Assessment of Hand Motion Analysis during Simulated Shoulder Arthroscopy.

PURPOSE: Attempts to quantify hand movements of surgeons during arthroscopic surgery faced limited progress beyond motion analysis of hands and/or instruments. Surrogate markers such as procedure time have been used. The dimensionless squared jerk (DSJ) is a measure of deliberate hand movements. This study tests the ability of DSJ to differentiate novice and expert surgeons (construct validity) whilst performing simulated arthroscopic shoulder surgical tasks. METHODS: Six residents (novice group) and six consultants (expert group) participated in this study. Participants performed three validated tasks sequentially under the same experimental setup (one performance). Each participant had ten performances assessed. Hand movements were recorded with optical tracking system. The DSJ, time taken, total path length, multiple measures of acceleration, and number of movements were recorded. RESULTS: There were significant differences between novices and experts when assessed using time, number of movements with average and minimal acceleration threshold, and DSJ. No significant differences were observed in maximum acceleration, total path length, and number of movements with 10m/s2 acceleration threshold. CONCLUSION: DSJ is an objective parameter that can differentiate novice and expert surgeons' simulated arthroscopic performances. We propose DSJ as an adjunct to more conventional parameters for arthroscopic surgery skills assessment.

Functional Classification of the Medial Ulnar Collateral Ligament: An In Vivo Kinematic Study With Computer-Aided Design.

BACKGROUND: It has been widely accepted that the anterior and posterior bundles of the medial ulnar collateral ligament (MUCL) tighten at extension and flexion, respectively. However, this belief is based on anatomic data acquired from cadaveric studies. The advancement of 3-dimensional (3D) model technology has made possible the simulation of dynamic movement that includes each ligament bundle fiber to analyze its functional properties. To date, no study has analyzed ligament kinematics at the level of the fibers while also focusing on their functional properties. PURPOSE: To propose a new classification for functional properties of the MUCL based on its kinematic pattern. STUDY DESIGN: Descriptive laboratory study. METHODS: Five healthy elbow joints were scanned by use of computed tomography, and 3D models were rendered and translated into vertex points for further mathematical analysis. The humeral origin and ulnar insertion of the MUCL fiber groups were registered. Each vertex point on the origin side was randomly connected to the insertion side, with each pair of corresponding points defined as 1 ligament fiber. Lengths of all the fibers were measured at 1° increments of elbow range of motion (ROM). Ligament fibers were grouped according to their patterns. Mean coverage area for each group, expressed as the percentage of ligament fibers per group to the total number of fibers, was calculated. RESULTS: Four major bundle groups were found based on fiber length properties. Kinematic simulation showed that each group had a different kinematic function throughout elbow ROM. Mean coverage area of groups 1, 2, 3, and 4 was 8% ± 4%, 10% ± 5%, 42% ± 6%, and 40% ± 8%, respectively. Each group acted as a dominant stabilizer in certain arcs of motion. Reciprocal activity was observed between groups 1 and 3 along with groups 2 and 4 to produce synergistic properties of maintaining elbow stability. CONCLUSION: Detailed analysis of fibers of the MUCL allows for further understanding of its kinematic function. This study provides MUCL group coverage area and kinematic function for each degree of motion arc, allowing selective reconstruction of the MUCL according to mechanism of injury. CLINICAL RELEVANCE: Understanding the dominant functional fibers of the MUCL will benefit surgeons attempting MUCL reconstruction and will enhance further anatomic study.

The Effect of Split Nerve on Electromyography Signal Pattern in a Rat Model.

BACKGROUND: Recent developments of prosthetic arm are based on the use of electromyography (EMG) signals. To provide improvements, such as coordinated movement of multiple joints and greater control intuitiveness, higher variability of EMG signals is needed. By splitting a nerve lengthwise, connecting each half to new target muscles, and employing a program to assign each biosignal pattern to a specific movement, we hope to enrich the number of biosignal sites on amputees' stump. METHODS: We split the gastrocnemius muscle of 12 Sprague-Dawley rats into two muscle heads, searched for the peroneal nerve, divided them lengthwise, and connected one half of the nerve to the tibial nerve innervating both muscle heads (SN_50, n = 8). In another group, we connected the undivided peroneal nerve to the nerve of a single muscle head (non-SN_100, n = 6), while the other muscle head received different innervation (non-SN_0, n = 6). After 10 weeks, we stimulated the peroneal nerve and measured the EMG amplitude. RESULTS: Mean EMG amplitude of the muscle head innervated by one half of the nerve (SN_50; 1.77 [range: 0.71-3.24] mV) and by the undivided nerve (non-SN_100; 3.45 mV [range: 1.13-5.34]) was not significantly different. However, the mean EMG amplitude produced by SN_50 was significantly different from that of the other innervation (i.e., non-SN_0; 0.76 mV [range: 0.41-1.35]), indicating the presence of noise. CONCLUSION: Split nerve in combination with split-muscle procedure can yield a meaningful EMG signal that might be used to convey the intention of living organism to a machine.

Isometric Tunnel Placement in Ulnar Collateral Ligament Reconstruction with Single CT Scan.

BACKGROUND: Isometric tunnel placement for anterior bundle of the medial collateral ligament (MCL) reconstruction is mandatory for successful surgery. PURPOSE: This study aimed to demonstrate a useful method for identifying isometric tunnel placement using a single computed tomography (CT) scan. STUDY DESIGN: Descriptive Laboratory Study. METHODS: Five normal elbows were scanned at 4 different flexion angles at 45° increment. Three-dimensional models were analyzed using 2 different approaches: single and multiple CT scans methods. Ligament footprints in the humerus and the ulna were registered. Ligament length and isometric points were defined. The locations of the isometric points were imported into both methods to be compared. RESULTS: There was no significant difference between 2 methods in calculating the length in every zone. There was also no significant difference in determining isometric ligament's origin point, which is located approximately 18.2 ± 4.0 mm and 18.4 ± 2.9 mm for single and multiple CT, respectively, measured inferolaterally from medial epicondyle. CONCLUSIONS: A solid preoperative plan is critical when predicting tunnel locations due to the difficulty in finding isometric points and the individuality of optimal bone tunnel locations. Using single CT scan, optimal locations can be predicted with the same accuracy as a multiple CT scans with less radiation exposure.

Beveled posteromedial corner of the radial head: a three-dimensional micro-computed tomography modeling study.

The posteromedial quadrant of the radial head is known to be different from the other quadrants. However, the explanation of this unique anatomical feature remains elusive. Hence, this study was designed to address this unique anatomical variance using three-dimensional µCT (micro-computed tomography) analysis. Nine fresh cadaveric radial heads were scanned using µCT. Three-dimensional subchondral bone and cartilage models were rendered. Both models were separated into the four quadrants at both the periphery (rim) and the articulating dish (fovea): anteromedial (AM), posteromedial (PM), posterolateral (PL), and anterolateral (AL). Each quadrant was analyzed in terms of (1) subchondral bone porosity (SBP), (2) mean subchondral bone thickness (MSBT), and (3) mean cartilage thickness (MCT). There was a significant difference between the fovea and the rim in terms of its microarchitectural features. Although within the fovea, the PM quadrant did not differ significantly from the other quadrants, a significant difference was found within the rim. In terms of SBP, PM, AM, PL and AL were calculated as 33, 37, 36 and 35%, respectively. In terms of MSBT, PM, AM, PL and AL were calculated as 0.11, 0.10, 0.09, and 0.09 mm, respectively. In terms of MCT, PM, AM, PL and AL were calculated 1.09, 0.81, 0.84 and 0.83 mm, respectively. The PM corner of the radial head between the 8 and 9 o'clock positions, was beveled. This might explain why the PM quadrant of the rim differed significantly from the other quadrants in terms of its microarchitectural features.

Digital data acquisition of shoulder range of motion and arm motion smoothness using Kinect v2.

BACKGROUND: Range of motion (ROM) is a clinically important parameter in evaluating joint function. However, dynamic evaluation to determine the quality of the arm motion using digitized measurement is often overlooked during clinical assessment. We evaluated the accuracy of Kinect v2 (Microsoft, Redmond, WA, USA) as a digital tool for measuring shoulder ROM objectively and proposed a concept of motion smoothness reflecting the quality of arm motion. METHODS: Ten male participants were included in a 2-stage experiment. First, shoulder ROM was measured in 4 static poses (flexion, abduction, external rotation, and internal rotation) with Kinect v2, a 3-dimensional (3D) motion analysis system, and goniometry. Second, participants performed a point-to-point arm motion as naturally as possible. Kinematic data were collected with Kinect v2 and the 3D motion analysis system and then postprocessed to acquire parameters related to motion smoothness, including peak to mean velocity ratio, acceleration to movement time ratio, and number of peaks. RESULTS: Kinect v2 resulted in very good agreement of ROM measurement (r > 0.9) with the 3D motion analysis (95% limits of agreement < ±8°) compared with goniometry (95% limits of agreement < ±10°). Kinect v2 also showed a good correlation and agreement of measurement of motion quality parameters compared with the 3D motion analysis (peak to mean velocity ratio, acceleration to movement time ratio, and number of peaks: r = 0.769, discrepancy = ±0.1; r = 0.922, discrepancy = ±5%; and mean = 1 ± 0, respectively). CONCLUSIONS: We show that Kinect v2 can be used as a reliable tool to measure shoulder ROM and arm motion smoothness.

A hands-free region-of-interest selection interface for solo surgery with a wide-angle endoscope: preclinical proof of concept.

BACKGROUND: A hands-free region-of-interest (ROI) selection interface is proposed for solo surgery using a wide-angle endoscope. A wide-angle endoscope provides images with a larger field of view than a conventional endoscope. With an appropriate selection interface for a ROI, surgeons can also obtain a detailed local view as if they moved a conventional endoscope in a specific position and direction. METHODS: To manipulate the endoscope without releasing the surgical instrument in hand, a mini-camera is attached to the instrument, and the images taken by the attached camera are analyzed. When a surgeon moves the instrument, the instrument orientation is calculated by an image processing. Surgeons can select the ROI with this instrument movement after switching from 'task mode' to 'selection mode.' The accelerated KAZE algorithm is used to track the features of the camera images once the instrument is moved. Both the wide-angle and detailed local views are displayed simultaneously, and a surgeon can move the local view area by moving the mini-camera attached to the surgical instrument. RESULTS: Local view selection for a solo surgery was performed without releasing the instrument. The accuracy of camera pose estimation was not significantly different between camera resolutions, but it was significantly different between background camera images with different numbers of features (P < 0.01). The success rate of ROI selection diminished as the number of separated regions increased. However, separated regions up to 12 with a region size of 160 × 160 pixels were selected with no failure. Surgical tasks on a phantom model and a cadaver were attempted to verify the feasibility in a clinical environment. CONCLUSIONS: Hands-free endoscope manipulation without releasing the instruments in hand was achieved. The proposed method requires only a small, low-cost camera and an image processing. The technique enables surgeons to perform solo surgeries without a camera assistant.

Reinnervated Split-Muscle Technique for Creating Additional Myoelectric Sites in an Animal Model.

BACKGROUND: This study proposes a novel reinnervated split-muscle operation to create additional myoelectric sites as sources of command signals of myoelectric prostheses for enhanced dexterous hand-to-wrist motions. The aim of this study was to investigate the postprocedure electromyographic properties of the muscles as distinct myoelectric sites in a rat model. METHODS: The reinnervated split-muscle group (n = 6) had the gastrocnemius muscle separated along its longitudinal axis and nerves transferred to each new muscle (peroneal nerve to lateral muscle head and tibial to medial one); the non-split-muscle group (n = 6) only had nerve transfers with its muscle intact. Functional testing was conducted after 10 weeks. The main parameter is the difference in mean electromyographic amplitude between the new muscles, with greater values indicating better separability. RESULTS: After the reinnervated split-muscle procedure, there is a significant increase of the average ratio between two muscles compared with the control group, from 0.44 (range, 0.02 to 0.86) to 0.77 (range, 0.35 to 0.98) (p = 0.011). In addition, compared with the non-split muscle group, nerve transfer in the split-muscle group is more successful in reaching its intended target muscle. CONCLUSION: A reinnervated split-muscle procedure could be beneficial for acquiring a more precise and discrete command signal in upper limb amputees, thus enabling the creation of more dexterous prosthetic arm.

Dorsal apex curve of the proximal ulna.

OBJECTIVE: To analyze the apexes and angulations along the curvature in the posterior border of the proximal ulna-termed the "dorsal apex curve" (DAC)-using the computer-aided design software. METHODS: Eight pairs of normal cadaveric ulnae were analyzed. The ulnar Cartesian coordinate system was standardized using a user-defined coordinate systems feature. DAC was defined by the best-fit curvature of serial apexes in the posterior border in axial cross-sections along the ulna, whereas the best-fit curvature was obtained using polynomials interpolation method. DAC apexes in three different planes were determined by calculating the second derivative value of curve function, and subsequently, DAC angulations were calculated. Statistical analysis was performed using analysis of variance with repeated measures with significance level set as 0.001. RESULTS: The average ulnar length was 246±15 mm (224-274 mm). The average ratios of dorsal apex, varus apex, and edge point were 23±3% (17-27%), 33±4% (27-43%), and 10±1% (8-13%), respectively, relative to the ulnar length. The average amount of varus, dorsal, hook, and torsion angulation were 167±4° (157-172°), 176±1° (175-178°), 90±0°, and 31±10° (15-49°), respectively. We found no significant differences between left and right ulnae or between male and female ulnae. CONCLUSION: Dorsal apexes and angulations are important landmarks for surgeons when applying plates and attempting total elbow replacement surgery. This curve provides valuable information to medical manufacturers for modeling both plates and ulna components of artificial elbow joints.

A new wide-angle arthroscopic system: a comparative study with a conventional 30° arthroscopic system.

PURPOSE: To compare users' hand movements in performing validated shoulder arthroscopic tasks between a 30° and a wide-angle arthroscopic system, using phantom models with an optical motion analysis system. METHODS: Twelve orthopaedic residents were enrolled and randomly allocated into two groups. In order to compensate for any learning effect, a Latin square counterbalancing technique was used. An optical motion analysis system was used with markers affixed to pre-designed sites; each participant conducted four validated shoulder arthroscopic tasks using both arthroscopic systems. Each participant was instructed to perform the experiment three times with each arthroscope. The time taken, total path length, number of movements, and average acceleration were analysed. RESULTS: Significant differences were observed for the time taken, number of movements, and average acceleration between the two arthroscopic systems (P < 0.05 for all). However, the time taken was not significant. The mean total path length measured 53 ± 38 cm with the 30° arthroscope, while the mean with the wide-angle arthroscope was significantly shorter, at 36 ± 22 cm. The mean number of movements with the 30° and wide-angle arthroscopes were 1974 ± 1305 and 1233 ± 990, respectively, while the average accelerations were 2.6 ± 1.3 and 1.2 ± 0.6 cm/s(2), respectively. The mean time taken was 13 % faster when using the wide-angle arthroscopic system, although this was not statistically significant. CONCLUSION: The wide-angle arthroscopic system improved the arthroscope manoeuvre in terms of the total path length, number of movements, and average acceleration required for experimental arthroscopy. This system may help surgeons triangulate the arthroscope and surgical instruments during surgery by expanding the field of view.

Comparative analysis of visual field and image distortion in 30° and 70° arthroscopes.

PURPOSE: There have been very few attempts to compare the visual fields and image distortion in arthroscopes. To better understand the images generated using existing arthroscopes, we performed image-mapping experiments to assess field of view and image distortion. The purpose of this study was to quantify and compare the visual fields obtained using 30° and 70° arthroscopes and assess image distortion in each arthroscope. METHODS: A complete arthroscopy system was used in this study. To perform this quantitative analysis, we created a customized measurement device that consisted of three parts: (1) distance marker, (2) chessboard pattern, and (3) angle marker. Three observers collectively assessed the appropriate position of the arthroscope during simulation. For each scope, ten simulations were performed at distances between 1 and 3 cm. Using the obtained arthroscope images, field of view and image distortion were measured and calculated. RESULTS: The field of view of the 70° arthroscope was 5, 10, and 15 mm wider in diameter in comparison with the 30° arthroscope at 1, 2, and 3 cm, respectively. Moreover, the 70° arthroscope had less 0.66, 0.13, and 0.26 pixels of root-mean-square distance than the 30° arthroscope at 1, 2, and 3 cm, respectively. The 70° arthroscope also contained 0.78 pixels less at the maximal error than the average 30° arthroscope. Therefore, the 70° arthroscope demonstrated less distortion than the 30° arthroscope. There was no significant difference between the two scopes with respect to median curvature measurement at 1-cm distance. CONCLUSION: The 70° arthroscope demonstrates technical advantages over the 30° arthroscope, including a wider field of view and a less image distortion at the periphery. A wide angle and less image distortion can help better orient the surgeon within the joint cavity when a panoramic picture is needed to repair rotator cuff tears, in case of hip arthroscopy, or while treating the lesions of posterior horn of medial meniscus.

Decellularized Nerves for Upper Limb Nerve Reconstruction: A Systematic Review of Functional Outcomes.

BACKGROUND: This is a systematic review for evaluating the evidence for functional outcomes after decellularized nerve use in clinical setting. Decellularized nerves are allografts whose antigenic components have been removed, leaving only a scaffold that promotes the full regeneration of axons. METHODS: Literature research was performed using the PubMed/MEDLINE database for English language studies with the keywords "decellularized nerve" and "processed nerve allograft." Inclusion criteria were prospective and retrospective case reviews in clinical settings. Exclusion criteria were case reports and case series. RESULTS: We retrieved six level VIII studies and one level VI study (classified according to the Jovell and Navarro-Rubio scale) with a total of 131 reconstructions. The basic data ranges of the studies were as follows: patient age, 18 to 86 years; duration between initial injury and nerve reconstruction procedure, 8 hours to 4 years; and follow-up period, 40 days to 2 years. The maximum lengths of the nerve gap for chemically washed decellularized nerves and cryopreserved decellularized nerves were 50 and 100 mm, respectively. Quantitatively, the functional outcome ranges were as follows: static two-point discrimination, 3 to 5 mm; and moving two-point discrimination, 2 to 15 mm. For motor assessment, all patients had a > M3 Medical Research Council score. It is also important to notice that a large variability occurs in almost every factor in the reviewed studies. CONCLUSION: Our study is the first to summarize the clinical results of decellularized nerves. Decellularized nerves have been used to bridge nerve gaps ranging from 5 to 100 mm with associated satisfactory outcomes in static and moving two-point discriminations.

Vortical flow in human elbow joints: a three-dimensional computed tomography modeling study.

The human elbow joint has been regarded as a loose hinge joint, with a unique helical motion of the axis during extension-flexion. This study was designed to identify the helical axis in the ulnohumeral joint during elbow extension-flexion by tracking the midpoint between the coronoid tip and the olecranon tip of the proximal ulna in a three-dimensional (3D) computed tomography (CT) image model. The elbows of four volunteers were CT-scanned at four flexion angles (0°, 45°, 90°, and 130°) at neutral rotation with a custom-made holding device to control any motion during scanning. Three-dimensional models of each elbow were reconstructed and a 3D ulnohumeral joint at 45°, 90°, and 130° was superimposed onto a fully extended joint (0°) by rotating and translating each 3D ulnohumeral joint along the axes. The midpoints of the olecranon and coronoid tips were interpolated using cubic spline technique and the dynamic elbow motion was plotted to determine the motion of the helical axis. The means and standard deviations were subsequently calculated. The average midpoint pattern of joint motion from extension to flexion was elliptical-orbit-like when projected onto a sagittal plane and continuously translated a mean 2.14 ± 0.34 mm (range, 1.83-2.52 mm) to the lateral side during elbow extension-flexion. In 3D space, the average midpoint pattern of the ulnohumeral joint resembles a vortical flow, spinning along an imaginary axis, with an inconsistent radius from 0° to 130° flexion. The ulnohumeral joint axis both rotates and translates during elbow extension-flexion, with a vortex-flow motion occurring during flexion in 3D model analysis. This motion should be considered when performing hinged external fixation, total elbow replacement and medial collateral ligament reconstruction surgery.

Morphometric analysis of the proximal ulna using three-dimensional computed tomography and computer-aided design: varus, dorsal, and torsion angulation.

PURPOSE: The proximal ulna, particularly the course of the posterior border, has a complex three-dimensional (3D) morphology which has been highlighted recently due to its clinical relevance in relation to surgical treatments. 3D computed tomography (CT) reconstruction and computer-aided design (CAD) based software can help to visualize the complex anatomy and thus aid the investigation of the more detailed morphology of the proximal ulna. METHODS: In our current study, 3D CT reconstruction images of 20 cadavers were imported into the 3D CAD program. Three morphologic angle parameters of the proximal ulna were measured including the dorsal, varus and torsion angulation. The torsion angulation was measured using the flat spot of olecranon dorsal aspect. We measured the total length of the ulna and the distance between the olecranon tip and the apex of dorsal and varus angulation. Furthermore, the thickness of olecranon was also measured for all the specimens. RESULTS: The results showed that the mean dorsal, varus, and torsion angulation was 4.3° (range 2.6°-5.9°), 12.1° (range 7.9°-17.6°), and 22.5° (range 16.6°-30.5°), respectively. The average length ratio of the dorsal and varus angulation apex to the total ulnar length was 26.4 % (range 19.8-30.7 %) and 32.7 % (range 27.5-37.5 %), respectively. The average of olecranon thickness at the proximal tip, mid-olecranon fossa, and at coronoid tip level was 17.8 mm (range 14.1-22.8 mm), 19.7 mm (range 15.8-23.1 mm), and 35.1 mm (range 27.9-41.8 mm), respectively. CONCLUSION: In conclusion, variations in the proximal ulna have to be considered when anatomically contoured dorsal plates are applied. Knowledge of the 3D morphologic anatomy of the proximal ulna would provide important information on fracture reductions, and the design of a precontoured dorsal plate or a prosthetic ulnar stem.