Worldwide 275,000 diseased heart valves are replaced annually and approximately 50% are bioprosthetic heart valves (BHV). BHV are fabricated from biologically derived tissues chemically modified to reduce immunogenecity and improve durability. BHV are nonviable, non-renewing biomaterials that undergo progressive degenerative changes in-vivo resulting in durability issues, which can be due to both calcific and non-calcific mechanisms. In-vitro durability testing of intact valves up to 200x106 cycles is used to assess BHV durability. In-vitro durability testing confounds characterization and modeling of fatigue. Thus there is a need for elucidation of the underlying mechanisms in the BHV response to repeated cyclic loading (RCL), independent of BHV design. In this study, the effects of collagen orientation on the medium-term (up to 50x106 cycles) BHV RCL response was investigated. Glutaraldehyde treated bovine pericardium were subjected to cyclic tensile loading to stress levels of 500±50 kPa at a frequency of 22 Hz. Two specimen groups were examined, with the preferred collagen fiber direction parallel (PD) and perpendicular (XD) to the direction of loading. Small angle light scattering (SALS) was used to assess the degree of fiber reorientation of the BHV collagenous network after 0 and 50x106 cycles. After 0, 20x106 and 50x106 cycles, specimens were subjected to biaxial mechanical testing and Fourier transform IR spectroscopy (FT-IR) was performed to assess molecular level changes to collagen . In addition, and the collagen fiber crimp period was also measured. Substantial permanent set effects were observed in both groups. In the perpendicular group, the areal stretch, which is a measure of overall tissue compliance, increased significantly while in the parallel group the areal stretch decreased significantly after 50x106 cycles. After 50x106 cycles, SALS measurements revealed that in the perpendicular group, the collagenous fibers became less aligned and in the parallel group, the collagen fibers became more highly aligned. The only significant changes in collagen crimp were an increase in collage crimp period from 23.46±1.39 mm (at 0 cycles) to 28.14±0.84 mm after 20x106 cycles in the parallel group. FT-IR spectra indicated that RCL of both of the groups lead to collagen conformational changes and early denaturation after 20x106 cycles. The results of this study suggest that 1) collagen orientation plays a critical role in BHV fatigue response, and 2) chemical fixation technologies that allow greater fiber mobility under functional stresses yet without permanent set effects may yield more durable materials.

Injuries to the acromioclavicular joint usually result in pain and instability. However, few biomechanical studies have investigated the mechanism of injury and treatment. Consequently, current rehabilitation protocols and surgical techniques have similar outcomes with no gold standard for treatment. Therefore, the thesis objective was to evaluate the function of the intact and injured acromioclavicular joint during combined loading to provide guidelines for the development of rehabilitation protocols and reconstructions for complete dislocations. A robotic/universal force-moment sensor testing system was utilized to apply an external load in combination with joint compression to intact and injured joint. Joint compression caused significantly decreases in primary (in the direction of loading) translations and joint contact forces while increasing coupled (orthogonal to the direction of loading) translations for the injured joint (p<0.05). These findings suggest common surgical techniques such as distal clavicle resection, which remove painful joint contact, may cause loads to be supported by other structures and be transmitted over a smaller area due to the increased coupled motion and joint contact. Both findings reinforce the importance of restoring each component of the acromioclavicular joint after injury. Next, the cyclic behavior and structural properties of an anatomic reconstruction of the coracoclavicular ligaments were determined during uni-axial tensile testing and compared to the intact coracoclavicular ligaments. After complete dislocation of the acromioclavicular joint, the anatomic reconstruction complex was found to have significantly lower stiffness and ultimate load compared to the intact ligaments (p<0.05). However, the bending stiffness of the clavicle significantly decreased after dislocation. Consequently, the individual properties of the tendon graft used during reconstruction had more comparable results to the intact coracoclavicular ligaments than current surgical techniques. The evaluation of the intact and injured joint during a combination of loading conditions provided guidelines for the development of an anatomic reconstruction. The experimental methodology used to evaluate the anatomic reconstruction incorporated a more realistic mechanism of failure scenario during testing. Both studies provide insight on functional changes of the intact acromioclavicular joint following injury and reconstruction. Future investigations should quantify the loads transmitted across the joint during daily activities. Computational models could use this information in addition to data collected with the methodology developed in this thesis to optimize the proposed anatomic reconstruction.

A non-invasive, non-contact methodology to estimate forces in the anterior cruciate ligament (ACL) in response to in vivo knee kinematics will allow ACL surgical procedures and rehabilitation protocols to be improved. The specific aim of this study is to evaluate the feasibility of a non-invasive, non-contact methodology for estimating force in the ACL by reproducing average differential kinematics in 6-degrees of freedom (DOF) from one set of porcine knees (source) onto a separate set of porcine knees (target). Differential kinematics are motions of the knee in 6-DOF relative to the passive path of flexion-extension 1. Differential kinematics from a set of source cadaveric porcine knees (n = 8) were recorded in response to an anterior load of 100 N and a valgus load of 5 Nm at 30°, 60°, and 90° of knee flexion. The in situ forces in the ACL of the target knees (n = 8) in response to reproducing average differential kinematics was compared to the in situ forces in the ACL of target knees resulting from the application of the same anterior and valgus loads. There was a significant difference in the in situ force in the ACL between applied loads and average differential kinematics for all flexion angles under anterior loading and at 60° of knee flexion for valgus loading. There was not a significant difference in the in situ force in the ACL for valgus loading at 30° or 90° of knee flexion. Under anterior loads, in situ force in the ACL from reproducing average differential kinematics and applied loads differed by up to 227% in two target knee; although, the anterior tibial translations were identical. These results indicate that average differential kinematics from a random sampling of knees does not account for the 6-(DOF) motion of the knee. This is because variations in knee laxity cause coupled motions to be averaged out of the differential kinematics, artificially constraining the knees motion. In the future, cadaveric knees will be matched to the group of subject kinematics with similar anterior and internal-external knee laxity to improve estimates of the forces in the ACL.

Delineation of the load-displacement characteristics of osteoligamentous spinal specimens has become fundamental to the investigation of spinal biomechanics. Traditionally, in-vitro kinetic parameters of the spine have been obtained through flexibility tests employing open or closed loop "load control" methods, or stiffness tests employing "displacement control" methods-each control method having attendant advantages and disadvantages. On the other hand, the combination load control and displacement control methods into a new, "hybrid control" method have advantages over load control or displacement control alone. Further, physical evidence such as presence of certain receptors suggests that the human body may employ a type of hybrid control method in the control of spinal movements. In the present study, a robotics-based spine testing system with hybrid control was developed to delineate the in-vitro kinetics of lumbar spine specimens. The testing system was validated experimentally using a physical rigid-body-spring model of a spine specimen, as well as analytically by computer simulations in Matlab. For systematic study, the two components making up a hybrid control algorithm were analyzed separately: the outer "displacement control" loop, and the inner "load control" loop. The outer loop applies a rotation (e.g., flexion/extension) to the specimen, while the inner loop minimizes unwanted coupled forces (e.g., anterior/posterior shear and axial tension/compression). The performance of existing standard hybrid control algorithms was tested in terms of a number of parameters, including peak force, work done to a specimen, and number of iterations. Based on these tests, a number of proposed changes to improve algorithm performance were identified. Updating the user-defined center of rotation (COR) to reflect a specimen's COR was found to improve performance of the displacement control part of the hybrid control algorithm, while using a more completely populated stiffness matrix improved performance of the load control part. The re-combination of the displacement control and load control loops into the fully constituted hybrid control algorithm revealed interesting interactions between these control components that suggest a basis for spinal dysfunction.

Although wheelchair users are constantly subjected to oscillatory and shock vibrations not much research has been conducted to assess the whole-body vibrations experienced by wheelchair users. Studies that have been published have only involved the testing of manual wheelchairs not interventions such as suspension or seating systems. The purpose of this study was to determine if selected wheelchair cushions reduce the amount of harmful whole-body vibrations transferred to wheelchair users and, if the absorbed power method a good measure of evaluating the whole-body vibrations. Thirty-two participants, who use a wheelchair as their primary mode of mobility, partook in this study. Four of the most commonly prescribed wheelchair cushions were selected. Participants were asked to propel their wheelchair over a simulated activities of daily living (ADL) obstacle course while acceleration and force data was collected. A repeated measures ANOVA showed no significant differences between the different cushions for the total averaged absorbed power (p = .190), the 50 mm curb drop (p = .234), or the rumble strip (p = .143). A repeated measures ANOVA for the peak curb drop absorbed power revealed a significant difference in the cushions (p = .043). The cushions that appeared to perform the best in this testing appear to be the Invacare Pindot and the Varilite Solo. Not only did those cushions appear to have the lowest values much of the time but did not display the highest values. Absorbed power appears to be just as effective at determining the effects of vibrations in the time domain as the prescribed methods of the ISO 2631 standard.

The Veterans Affairs Personal Adaptive Mobility Aid (VA-PAMAID) is a robotic walker that is designed to provide physical support and obstacle avoidance and navigational assistance to frail visually impaired individuals. The goal of this study was to develop and implement testing protocols to determine the performance and safety capabilities of the device and use the results to redesign the walker to make it more reliable and effective. Engineering tests were performed to determine factors such as stability, range, speed, and fatigue strength. Additional tests to characterize the reliability and accuracy of the sensors and avoidance/navigation algorithms were also conducted. The walker traveled 10.9 kilometers on a full charge, and was able to avoid obstacles while traveling at a speed of up to 1.2 m/s. There were no failures during static stability, climatic, or static, impact, and fatigue testing. Some problems were encountered during obstacle climbing and sensor and control testing. Several significant differences were found with respect to the detection distance of the device when varying the obstacle height, material, approach angle, and lighting source. The walker also failed to detect 40-50% of the doorways during the hallway test. Clinical trials were conducted to compare the VA-PAMAID to a low-tech mobility aid (AMD). Subjects were recruited and trained to use both devices efficiently. Each participant was then asked to traverse an obstacle course several times. The time to complete the course, number of wall and obstacle collisions, and number of reorientations were all recorded and averaged. There were no significant differences between the VA-PAMAID and the AMD with respect to collisions or reorientations. The AMD had a significantly lower completion time (p=0.017) than the VA-PAMAID on the obstacle course. The results of the engineering and clinical tests were then used in a house of quality model to determine what factors of the walker needed to be revised. Specific modifications were recommended that would make the device safer, more reliable, and more marketable. Changing the wheel size, mass, component positions, detection algorithm, and other variables would make the VA-PAMAID easier to use and more effective for elderly visually impaired individuals.

Injury to the anterior cruciate ligament (ACL) of the knee can result in joint instability even following reconstruction. This instability may be quantified by measuring in vivo knee kinematics in six degrees of freedom. Motion capture systems have been used for measuring kinematics but they are limited by system inaccuracies and error resulting from skin movement. Therefore, the overall objective of this thesis was to determine the level of accuracy that a motion capture system using surface markers provides for measuring knee kinematics. There are three specific aims in this thesis. The first specific aim was to mathematically investigate the effect of random errors in marker locations on the accuracy of knee kinematics calculated using the point cluster technique (PCT), triad and Helen Hayes marker sets. The results indicated that the PCT marker set had the greatest potential for accurately measuring knee kinematics. The second aim was to determine how inaccuracies of the motion capture system contribute to errors of joint kinematics measured with the PCT. Despite its high accuracy, the average errors of joint kinematics attributed to the system were up to 1° and 2 mm. The final specific aim was to investigate the efficacy of an algorithm called the interval deformation technique (IDT) for reducing errors of knee kinematics resulting from skin movement. The IDT reduced the errors of kinematics by 90% for an activity with skin movement simulating muscle contraction but was unable to reduce the errors resulting from skin movement at heel strike of gait. The overall errors of knee kinematics resulting from system inaccuracies and skin movement were estimated to be 2° and 4 mm. While this technique may be useful for measuring the changes in knee kinematics that result from ACL injury, this accuracy may not be sufficient to discern the small differences in knee kinematics between ACL intact and reconstructed subjects or for predicting ligament forces. Thus, further research is suggested in order to better quantify skin movement and provide data for improving the accuracy of kinematics measured with surface markers.

Total shoulder arthroplasty restores function in shoulders with end stage glenohumeral arthritis. The most common complication of arthroplasty is prosthesis loosening and glenoid prosthesis loosening occurs more frequently than humeral because of quantity and orientation of bone available for fixation. Increasing glenoid prosthesis longevity requires thorough understanding of scapula structure, especially glenoid morphology and bone density. The project aim was quantification of glenoid structure with specific relevance to improve arthroplasty. Detailed knowledge of glenoids intra-articular geometry, subchondral structure, regional bone density and extra-articular relationships is needed for future prosthesis design optimization. Three-dimensional computer models were generated from CT images of 12 pairs of male cadaver scapulae aged 50.18 ± 11.77 years, and 8 pairs of female cadaver scapulae aged 60 ± 20.48 years. External glenoid morphological parameters measured included superior-inferior length, anterior-posterior width, and glenoid contour geometry (dimensions and angles). Internal morphological analysis included subchondral bone glenoid version measurement. Regional bone density measurements were made to determine glenoid cancellous bone distribution. Accuracy and reliability were defined using repeated measurements. The glenoid was pear shaped with superior-inferior length greater than anterior-posterior diameter. The inferior glenoid boundary was a 120° arc with average radius 11.2 ± 1.2 mm. The center of the arc (glenoid center) was located along the maximum superior-inferior length one-third this distance superior from infraglenoid tubercle and. Glenoids articular surface version, and subchondral bone version averaged 2° ± 5°, and 1° ± 4° of retroversion, respectively. Highest density bone was in posterior glenoid, medium density anteriorly, and low density in central glenoid. Accuracy and reliability were defined as mean difference between repeated and original computer model measurement (0.5 ± 0.7 mm for lengths and 1.3° ± 4.4° for angles). 3-D computer modeling permitted internal morphological analyses, which for the first time defined entire glenoid structure. External morphological and bone density measurements agreed with previously reported data. Advanced imaging and computer modeling tools enabled an accurate and reliable structural analysis of the complexly shaped glenoid. Work described in this project will be used for future studies whose goals are improved glenoid prosthesis and surgical instrumentation design.

EFFECTS OF DIABETES MELLITUS ON THE BIOMECHANICAL PROPERTIES AND PHARMACOLOGICAL FUNCTION OF THE FEMALE RAT URETHRA EX-VIVO Rachelle Lynn Prantil, MS University of Pittsburgh, 2004 Diabetic cystopathy results in a grossly distended, hypomotile bladder due to inefficient voiding. While the bladder has been extensively studied, little effort has been made towards the understanding of the urethra and the effects of this devastating disease. The current study is aimed to evaluate the effects of diabetes mellitus (DM) on the biomechanical properties and the pharmacological function of the female rat urethra ex vivo. DM was induced in female rats by injection of streptozotocin. At 3, 5,and 10 weeks, the urethras were excised and mounted into an ex-vivo system at in vivo length. For mechanical testing, urethras were subjected to stepwise increases of static, intraurethral pressure from 0 to 20 mmHg in both a baseline and passive state. Continuous outer diameter measurements were made using a laser micrometer at proximal, middle, and distal portions of the urethra. Compliance and beta stiffness were calculated from measured data. Pharmacological experiments involved assessments of mid urethral outer diameter response to Nç Nitro-L-arginine, phenylephrine, and EDTA. Age matched normal urethras served as controls. Statistical comparisons were made via ANOVA. Tissue was then processed for immuno- and histochemical quantification of smooth muscle, collagen, and elastin. For baseline healthy tissue, results showed a proximal to distal compliance gradient (proximal most compliant and distal least compliant), and the passive state enhanced the observation. Baseline beta stiffness values showed an increased stiffness in proximal and middle urethral portions by 5 and 10 weeks, and baseline compliance values showed at low pressures showed an increase in proximal compliance at 3 weeks and a decrease in proximal compliance at 5 weeks at high pressures. Passive beta stiffness and compliance values indicated proximal urethral stiffening by 10 weeks DM. Pharmacological studies revealed that DM abolishes endogenous nitric oxide release and increases the time to reach maximal relaxation. In cases of severe DM, alpha 1 adrenergic contraction was minimized. Little or no differences were found in the amount collagen, smooth muscle, and elastin. From these findings, it can be concluded that DM causes urethral stiffening and impaired contractile and relaxation urethral mechanism. Damaged urethral properties and function have serious implications for outlet resistance; thus, contributing to diabetic cystopathy.

A major challenge facing the development of implantable devices for clinical use is in finding a suitable power source for such devices. The power source should be able to generate an electric current for extended periods of time. Biofuel cells (BFC) provide some promise in this respect, as their function is primarily based on coupling the oxidation of glucose to the reduction of molecular oxygen to water. Under ideal conditions, the only byproducts of the BFC would simply be carbon dioxide and water. Both glucose and oxygen are present in the cells and tissues of all eukaryotic organisms, including human beings. It might, therefore, be possible to tap into the body's own resources, including the metabolic properties of our cells, to generate enough energy to power an array of clinical devices. The experiments described in this paper serve as a first step toward the goal of designing a BFC that would be based on transducing the power of oxidative metabolism within our own cells into an electrical current. In the first phase of our experiments, the function and current output of a specific type of BFC, called a microbial fuel cell (MFC)is investigated. The behavior and characteristics of such biofuel cells have been well documented in the scientific literature. MFCs essentially convert the biochemical energy of bacteria into electrical energy. A strain of E. Coli is used in our study. In the second phase of our experiments, an attempt is made to derive electrical currents from BFCs employing human white blood cells.