Research : Biomechanics

Biomechanics Research Projects


Our faculty surgeons: Frank Eismont, Jonathan Gottlieb, Joseph Gjolaj
Our voluntary faculty: Harry Shufflebarger, Dan Cohen
Our UM engineers: Loren Latta, Edward Milne, David Kaimrajh, Francesco Travascio, Shihab Asfour
Our researcher: Ali Alhandi
Outside surgeons: Peter Ostermann, Ronald Lindsey, Robert Norton, Seth K. Williams
Medical students: Amir Qureshi, Mark Barton, Anita Vijapura
Engineer student: Jason Albert

Bone anchoring: In osteoporotic vertebral bodies a significant increase in strength of fixation and reduction of loosening under cyclic load with pedicle screws was achieved with a small amount of bone cement injected through the screw and into the threads.*

A strong correlation was found in cadaver spines between decreased bone mineral density and lower screw insertion torque in the vertebral endplates when applying anterior lumbar cages for fusion.*

Cadaver lumbar vertebral bodies had pedicle screws applied with bone cement injected through the screws on one side of the vertebral body, and not in the other side. A small Harms cage was then placed on each side of the superior end plate and the load to cause penetration of the end plate was measured. The same test was repeated on the inferior end plate. In normal vertebral bodies there was no difference in resistance to penetration with or without bone cement, but there was a significant difference in osteoporotic vertebral bodies. The inferior end plates were significantly stronger than the superior end plates for this test.*

Cadaver lumbar vertebral bodies were DEXA scanned to separate “osteopenic” from “normal” groups. A pedicle screw was applied to one side, and a cortical screw applied to the other. Rods were attached to each screw, and to an UHMWPE block (a la ASTM/ISO standardized test) and was loaded from 5 Nm extension, to 5 Nm flexion at 0.25 Hz. for 5,000 cycles and then subjected to screw pull-out test. Screw loosening was monitored by video fluoroscopy using ImagePro software. Loosening occurred early for both screws in the “osteopenic” group, but not in the “normal” group. Screw pull-out strengths for loose pedicle screws was significantly stronger than for loose cortical screws. However, pull-out strength for cortical screws that had not loosened was better than for pedicle screws that had not loosened.+

Fixation of L1 fractures: In a short segment construct of an L1 compression fracture simulated in cadaver spines, the addition of pedicle screw fixation at the L1 level significantly improved the rigidity of the construct.*

Compression of the facet joints within a short segment construct to stabilize a flexion-distraction injury, significantly increased the rigidity of the construct and reduced the loading in the rods in a cadaver model.* The same construct in a corpectomy model was improved, but not significantly, by adding the screws at L1.+

Scoliosis:. An existing synthetic model for testing long thoracic spinal constructs was validated for flexion-extension loading by comparison to human cadaver spines. Constructs from T1 – T10 were then cyclic tested with the synthetic model to evaluate potential loosening of components and wear debris.*

Cadaver spines from T12 – T1 were instrumented with 2 typical configurations for correction of scoliosis with an apex at T5. One configuration used hooks; the other, pedicle screws. The torsional rigidity of each construct was measured both with and without anterior fusion, simulated with bone cement in the intervertebral disc space. The rigidity of the pedicle screw construct without anterior fusion was higher than that of the hook construct with anterior fusion. The implication is that it may be possible to prevent “crank shaft” deformity from occurring with the use of pedicle screw constructs without the need for anterior fusion.*

Cervical spine: Cadaver cervical spines had motions at C6-7 and C3-4 measured before and after single and double level fusions at C4-5 and C5-6. There were significant changes in motion during flexion and extension loading on the whole spine, related to both single level fusions, but more dramatically in double level fusions.*

*These projects have completed the experimental work and were presented at national and international meetings in 2011. Abstracts of each have been published in 2011 and many are in the process of submission for publication scientific journals.

+These projects are ongoing and have more experimental work to be done in 2013 to complete them.


Our faculty surgeons: Stephen Quinnan, Gregory Zych
Our resident surgeons: Amar Patel
Our UM engineers: Loren Latta, Edward Milne, David Kaimrajh, Francesco Travascio, Shihab Asfour, Moataz Eltoukhy
Our researcher: Ali Alhandi
Outside surgeon: Ronald Lindsey
Medical students: Sara Pastoriza, Richard Morgan, Andrew Johnson, Evan Trapana, Jeff Shub
Engineer student: IEN Engr.s
Other students: Jordan Namer, Daniel Briggi, several volunteer subjects

Fracture fixation: Cadaver arms instrumented with 2 mm plates on the proximal radius showed no impingement with prono-supination, but 2.4 mm and 3.5 mm plate did impinge. There was also no statistical difference in bending stiffness in 3 planes or failure torque between the 3.5 mm plate and 2 mm 90-90 plating construct.+

The mechanical roles of a titanium mesh cage and an intramedullary nail in the reconstruction of significant segmental defects of the tibia was investigated invitro. Estimates of moments transmitted by the cage and the nail, show almost equal load sharing for the moments applied in valgus, anterior bowing and recurvatum. Only in varus loading did the cage carry most of the moment transmitted across the defect. The axial load transmission was carried primarily by the cage for all loading conditions.*

A non-bridging external fixation configuration was used to stabilize a 5 part, intra-articular fracture of the proximal humerus and humeral head. The construct was loaded in varus within the physiological range for 50,000 cycles. There was no measurable loss of rigidity of the construct or position of the bone fragments after cycling. The pins were each pulled from the pin holders and the pull-out strength of the pin-holder was at least as good as testing of freshly applied pins. So no loosening could be detected.*

Cadaver limb segments have a mid-diaphyseal, oblique fracture simulated by osteotomy in all long bones of the limb segment. In the thighs and upper arms, the femur and humerus are fixed. In the legs and forearms the tibias and radii are fixed. Fixation groups include plate and screws, intramedullary nails and single half-pin external fixators. The axial rigidity of each limb after fracture and fixation is measured under physiological load levels, then a splint or brace is applied with soft tissue compression in the area of the fixed fracture. This procedure is repeated after removing ¼ of the soft tissue circumferentially over ½ the length of the lateral side of each limb segment, and then again after removing 1/3 of the lateral soft tissue. Soft tissue compression with a brace does help to stabilize the limbs with intramedullary nails and those with external fixators, but not the limbs with plate and screw fixation. The role of soft tissue compression diminishes with loss of soft tissue. +

Pelvic fixation: Movements at the sacroiliac joint measured in human cadavers with flexion/extension, torsion and axial compression loading were altered by progressive fusion of L4-5 and then L5-S1.*

CT was used to map BMD regions in the sacrum to optimize the placement of screws for SI fixation. A single, large screw was compared biomechanically to existing screw fixation of the SI joint.* A model of the pelvis with plating of the SP joint showed motions at the SP joint were only slightly different between groups and none were statistically significant, though the vertical translations and the ML plane rotations, with the hybrid fixation were slightly reduced. The same was true for the SI joint vertical translations, but there was a significant improvement with the Hybrid fixation in ML plane rotations, P < 0.02. The closing of the SP joint with lateral load was significantly improved with the hybrid fixation, (P < 0.05). This study showed no advantage for locked plating.*

Distal radius fractures: Matched pairs of cadaver arms had a 5-part, intra-articular fracture of the distal radius fixed with a volar plate on one side, a non-bridging external fixation of the other. Bone fragment movements were recorded by fluoroscopy during loading in flexion, extension, ulnar deviation and radial deviation. The anatomic reduction of the externally fixed side was better than the volar plated side, and was maintained better under load. The plated side was significantly stronger than the externally fixed side.*

Using a simulated ‘fall on an outstretched hand’ model using 3 different hand positions to create a distal radius fracture we found that an ulnarly deviated or neutral wrist position resulted in a displaced distal radius fracture with a very strong linear correlation (0.84 and 0.87 respectively) between BMD and the force needed to create that fracture.+
Using a simulated ‘fall on an outstretched hand’ model using 3 different hand positions to create a distal radius fracture we found that that LT tears are more likely to be associated with fractures of the distal radius with internal forearm rotation and a radially deviated hand position.+

Using a simulated ‘fall on an outstretched hand’ model using 3 different hand positions to create a distal radius fracture we found that an ulnarly deviated wrist position was associated with a new scapholunate interosseous ligament tear in 2 out of 7 cadaveric arms compared to zero in 16 radially deviated or neutral arms.+

*These projects have completed the experimental work and were presented at national and international meetings in 2011. Abstracts of each have been published in 2011 and many are in the process of submission for publication scientific journals.

+These projects are ongoing and have more experimental work to be done in 2013 to complete them. 


Our resident surgeons: Razvan Nicolescu, Amar Patel
Our UM engineers: Loren Latta, Edward Milne, David Kaimrajh
Our researcher: Ali Alhandi
Outside surgeons: E. Anne Ouellette, Check Kam, Prasad Sawardeker, Chris Dy
Medical students: Ian Elliott, Tahsin Khundkar, Arianna Trionfo, Evan Trapana
Engineer student: Winston Elliott
Other student: Jordan Namer

Hand fractures: In human cadaver metacarpals, the multi-plane fixation of the extra-articular fragments provided by the NBX fixator was biomechanically superior to the fixation with 2 parallel pins in the Mini Hoffman fixator *

In human cadaver metacarpals, multi-plane fixation of the intra-articular fragments with the NBX fixator was biomechanically superior to the 2 parallel pins deployed by the Mini Hoffman fixator.*

Wrist injuries: The triquetral rotational index (TRI) is evaluated as an image analysis technique for diagnosis of ulnocarpal wrist instability in 105 healthy volunteers and 120 fresh human cadaver arms.*

Pathology of the TFCC is common cause of ulnar wrist pain but may also present with instability. In a cadaver study, we have demonstrated decreased stiffness in the UC joint after a peripheral-sided TFCC tear.*

Ulnar triquetral resistance to movement was measured in cadaver forearms before and after simulated TFCC tear, after stabilization with a functional brace designed for fracture treatment, and then again after surgical repair of the TFCC. The brace provided significant support to the injured TFCC.*

With supination tests of 30 cadaver arms, the examiner rated 4 wrists as stable and 26 as unstable. In biomechanical testing, the wrists judged to be stable had an average ulno-pisiform motion of 2.0 mm ±0.54 and wrists judged unstable had an average motion of 3.58 mm ±1.26. There was a statistically significant difference between the motion of the “stable” vs. “unstable” wrists.*

Cyclic loads applied to the flexor and extensor tendons in cadaver forearms drive the wrist motion while recording the positions of the carpal bones with fluoroscopy. The loading cycles are recorded for the intact wrist, after cutting the ligaments between the scafoid and lunate bones and then after repair of the simulated injury. These measurements were made with the wrist aligned in neutral, 20° radial deviation and then 20° ulnar deviation.+

Forearm axial injuries: Both lateral ulnotrochlear joint space and distal ulnar deviation showed significant increases corresponding to radial head implant sizing ranging in 2mm length increments, from -2mm to +4mm in cadaver arms.*

Measurements in cadaver arms showed a linear relation between RC and applied axial load, thus a consistent distribution of RC/UT at any load. Equivalence of ∆L values indicate consistent distribution with implant oversizing, showing constant interosseus ligament tension, which points to minimal distal radial translation.*

The percentage difference between radial head implant and capitellar radius of curvatures was plotted against corresponding measured pressures and contact areas in cadavers, showing a positive, linear trend.*

Successful results following radial head arthroplasty are dependent upon proper component position and size. In this cadaveric model progressive radial head lengthening was associated with a stepwise increase in radiocapitellar joint force. Radiocapitellar forces were 1.5 times greater with radial head overstuffing (+4) compared to neutral (0). Sizing of the radial head implant plays a critical role in restoring native radiocapitellar loads and may prevent accelerated wear at the radiocapitellar joint after radial head arthroplasty.*

Few biomechanical studies compare bipolar components to their monopolar counterparts. We examined the force conveyed at the radiocapitellar joint in a head-to-head comparison of bipolar impants to rigid, monopolar implants and found that pressures were notably higher when using monopolar radial heads.*

Axial load transfer was measured in intact cadaver arms and following complete disruption of the interosseous ligament complex, radial head excision and disruption of the DRUJ. Post injury, radial head arthroplasty improved radioulnar axial displacement but not to intact levels, doubled the radiocapitellar pressure and reduced the distal ulnar load by half that of intact arms.*

Axial loading on cadaver forearms with simulated Essex-Lopresti injuries showed that with radial head arthroplasty alone is sufficient to restore near normal biomechanics in partial disruptions of the IOLC. In complete IOLC injuries, radial head replacement alone doubled the radiocapitellar pressure.*

From cadaver studies of the forearm with Essex-Lopresti injury simulation and reconstruction, a method was devised to identify certain landmarks on the plain film X-rays that could be used to accurately determine the proper length of the radial head implant that would restore the length of the forearm. A triangle can be indentified with consistency that must be restored to create the proper length of the forearm.*

Measures of distal ulnar load transfer, radiocapitellar loading and motion of the distal radius and ulna to the carpal row with axial loading were used to monitor changes associated with removal of the radial head, then disruption of the DRUJ, followed by stripping of the interosseous ligament. After complete disruption the measures were repeated. The forearm was then progressively reconstructed by first applying a radial head implant, then repairing the interosseous ligament with either a patellar tendon graft or a tight rope. Next the DRUJ was stabilized with the Herbert sling procedure. The radial head alone did not restore normal load transfer to the forearm. The addition of the patellar tendon or the tight rope made a significant improvement, but still did not control the splaying of the radius and ulna. With the repair of the DRUJ, the normal mechanics of the forearm were almost completely restored.*

X-ray measures in cadaver forearms of the angles of the bones and the gaps in the joints at the elbow and wrist with radial head implants purposely placed at – 2 mm, 0 mm, +2 mm and +4 mm the normal length, showed significant changes in joint alignment and shift of the balance of loading from the radius and ulna at both the elbow and wrist.*

Fresh human forearms with all tissues intact were tested under axial loads with the forearm in neutral position, supination and pronation while the elbow was flexed to 90°and wrist was flexed 20°. The loading and axial stiffness measures were repeated after each of the following changes: 1. transducers placed in the radiocapitellar joint; 2. Transducers placed on the distal ulna; 3. Transducers placed on the interosseous ligament; 4. The skin and muscle tissues were stripped from the arm; 5. Large strain indicators were attached to the distal ulna and proximal radius. The objective was to detect any artifacts that might occur with the typical methods reported in the literature for analyzing the load transfer in the forearm under axial load. Many discrepancies exist in the literature. The hypothesis is that the methods deployed may account for the differences.+

*These projects have completed the experimental work and were presented at national and international meetings in 2011. Abstracts of each have been published in 2011 and many are in the process of submission for publication scientific journals.

+These projects are ongoing and have more experimental work to be done in 2013 to complete them.