Search

An EMG-driven musculoskeletal model is implemented to estimate subject-specific musculoskeletal parameters such as the optimal physiological muscle length, the tendon slack length and the maximum isometric muscle force of flexor and extensor muscle groups crossing the wrist, as well as biomechanical indexes to quantify the muscle operating range, the stiffness of the musculotendon actuators, and the contribution of the muscle fibres to the joint moment. Twelve healthy subjects (11 males and 1 female, mean age 31.1±8.7 years) were instructed to perform isometric maximum voluntary contractions of wrist flexors and extensors. Recorded EMGs were used as input to the model and the root mean square error (RMSE) between measured and predicted torque was minimised to estimate the subject-specific musculotendon parameters. The model was validated and the RMSE and the normalised RMSE calculated during estimation and validation phases are compared. Estimated subject-specific musculoskeletal parameters vary in a physiological range, while the biomechanical indexes are in agreement with previously published data. The proposed methodology proved to be effective for the in vivo estimation of physiological parameters of the musculotendon complex and has potential as an investigative tool to distinguish aetiological differences among subjects affected by musculoskeletal disorders., (Copyright © 2011 IPEM. Published by Elsevier Ltd. All rights reserved.)

An experimental and numerical analysis was performed for the Wang-Zwische double-lumen cannula (DLC) (Avalon Elite). The aim of this work was to provide insight for future improvement by characterizing the fluid dynamic behavior of the novel catheter with metrics often associated with blood trauma. Pressure and flow distributions were measured on a steady-flow rig using a 50% glycerol-water mixture by imposing a 2 L/min flow rate across the drainage and infusion lumens. The fluid was modeled as Newtonian with density of 1050 kg/m³ and dynamic viscosity of 0.0035 kg/m·s. Reynolds numbers typical for transitional flow (Re = 2000-2500) were computed within the lumens because of the changing cross-sections of the cannula geometry. Numerical computations were performed using the steady three-dimensional Reynolds-averaged Navier-Stokes (RANS) equations and the low-Reynolds k-ω turbulence model. Discretization of governing equations was based on a cell-centered finite volume method. Numerical results correlated well with global performance of the cannula, allowing evaluation of the geometry toward potential blood trauma. Peak wall shear stress (WSS) in the drainage lumen was higher than that of infusion lumen, mainly due to the presence of side holes. Furthermore, recirculation regions were predicted in transition tubing to connectors of both the drainage and the infusion lumens because of adverse pressure gradients caused by the sudden enlargement of the cannula geometry. In this three-dimensional computational fluid dynamics (CFD) study, we observed higher peak WSS values for the drainage lumen, which may potentially cause blood trauma. Furthermore, recirculation regions were predicted in the proximity of the exit sections of both the infusion and drainage lumens, which may contribute to thrombosis formation. This study provides insight for future DLC modifications in minimizing cannula-induced blood trauma and thrombogenicity in long-term applications.

Ventricular assist devices (VADs) are increasingly used for supporting blood circulation in heart failure patients. To protect or even to restore the myocardial function, a defined loading of the ventricle for training would be important. Therefore, a VAD control strategy was developed that provides an explicitly definable loading condition for the failing ventricle. A mathematical model of the cardiovascular system with an axial flow VAD was used to test the control strategy in the presence of a failing left ventricle, slight physical activity, and a recovering scenario. Furthermore, the proposed control strategy was compared to a conventional constant speed mode during hemodynamic changes (reduced venous return and arterial vasoconstriction). The physiological benefit of the control strategy was manifested by a large increase in the ventricular Frank-Starling reserve and by restoration of normal hemodynamics (5.1 L/min cardiac output at a left atrial pressure of 10 mmHg vs. 4.2 L/min at 21 mmHg in the unassisted case). The control strategy automatically reduced the pump speed in response to reduced venous return and kept the pump flow independent of the vasoconstriction condition. Most importantly, the ventricular load was kept stable within 1%, compared to a change of 75% for the constant speed. As a key feature, the proposed control strategy provides a defined and adjustable load to the failing ventricle by an automatic regulation of the VAD speed and allows a controlled training of the myocardium. This, in turn, may represent a potential additional tool to increase the number of patients showing recovery., (© 2010, Copyright the Authors. Journal compilation © 2010, International Center for Artificial Organs and Transplantation and Wiley Periodicals, Inc.)

A musculoskeletal model of wrist flexors comprising musculoskeletal dynamics and limb anatomy was experimentally validated with healthy subjects during maximum voluntary contractions. Electromyography signals recorded from flexors were used as input, while measured torques exerted by the hand were compared to the torques predicted by the model. The root mean square error and the normalized root mean square error calculated during estimation and validation phases were compared. In total, six subject-specific musculoskeletal parameters were estimated, while biomechanical indexes such as the operating range of the flexors, the stiffness of the wrist flexion musculotendon actuators, and the contribution of the muscle fibers to the joint moment were computed. Results are in agreement with previously published data.

This article describes an elastance-based mock ventricle able to reproduce the correct ventricular pressure-volume relationship and its correct interaction with the hydraulic circuit connected to it. A real-time control of the mock ventricle was obtained by a new left ventricular mathematical model including resistive and inductive terms added to the classical Suga-Sagawa elastance model throughout the whole cardiac cycle. A valved piston pump was used to mimic the left ventricle. The pressure measured into the pump chamber was fed back into the mathematical model and the calculated reference left ventricular volume was used to drive the piston. Results show that the classical model is very sensitive to pressure disturbances, especially during the filling phase, while the modified model is able to filter out the oscillations thus eliminating their detrimental effects. The presented model is thus suitable to control mock ventricles in real-time, where sudden pressure disturbances represent a key issue and are not negligible. This real-time controlled mock ventricle is able to reproduce the elastance mechanism of a natural ventricle by mimicking its preload (mean atrial pressure) and afterload (mean aortic pressure) sensitivity, i.e., the Starling law. Therefore, it can be used for designing and testing cardiovascular prostheses due to its capability to reproduce the correct ventricle-vascular system interaction.

During open-heart surgery roller pumps are often used to keep the circulation of blood through the patient body. They present numerous key features, but they suffer from several limitations: (a) they normally deliver uncontrolled pulsatile inlet and outlet pressure; (b) blood damage appears to be more than that encountered with centrifugal pumps. A lumped parameter mathematical model of a roller pump (Sarns 7000, Terumo CVS, Ann Arbor, MI, USA) was developed to dynamically simulate pressures at the pump inlet and outlet in order to clarify the uncontrolled pulsation mechanism. Inlet and outlet pressures obtained by the mathematical model have been compared with those measured in various operating conditions: different rollers' rotating speed, different tube occlusion rates, and different clamping degree at the pump inlet and outlet. Model results agree with measured pressure waveforms, whose oscillations are generated by the tube compression/release mechanism during the rollers' engaging and disengaging phases. Average Euclidean Error (AEE) was 20mmHg and 33mmHg for inlet and outlet pressure estimates, respectively. The normalized AEE never exceeded 0.16. The developed model can be exploited for designing roller pumps with improved performances aimed at reducing the undesired pressure pulsation.

In this work a pneumatically driven mock left ventricle and a native left ventricle are modeled and alternatively connected to the numerical model of a closed circulatory system comprising the systemic circulation, the left atrium, and inlet and outlet ventricular valves. By simulating physiological changes of the system working conditions, behavior and preload sensitivity of the pneumatic ventricle have been compared to those of a native ventricle. Results show that a pneumatic ventricle, when used as a fluid actuator in mock circulations, has low flexibility in reproducing different scenarios, its interaction with peripheral circulatory districts is characterized by non-physiological values, and its preload sensitivity is in poor agreement with physiological data. Results' analysis also shows that present mock circulatory systems for testing cardiovascular prostheses are inadequate, if a careful attention is not paid to the pumping action of the pneumatic ventricle. The presented computer model, validated by comparing numerical results with in vitro measurements available in the literature, can be used for designing in vitro experiments, while choosing the best control strategy for pneumatic systems.

The aim of this work is to investigate the dependence between left ventricular load impedance control by an apical ventricular assist device (VAD) and the consequent benefits for pathological heart recovery. A pathological left ventricle with 34% contractility has been simulated in the assisted and nonassisted conditions. By means of an extended Kalman filter, left ventricular pressure-volume loops have been partially estimated and ventricular as well as circulatory quantities inferred. The heart operation mode, based on cardiac energetic criteria, is imposed by controlling the VAD filling phase. In the assisted condition, results show that the left ventricle end-diastolic volume, left atrial pressure, and wall stress all decrease; stroke volume, ejection fraction, ventricular efficiency, aortic pressure, and cardiac output all increase. Benefits are also evident for the right ventricle and systemic and pulmonary circulation. The strategy outlined in this work also shows that good results for heart recovery are achievable and a possible way to improve the functional properties of commercial pulsatile VADs.

Hydraulic mock circulatory systems have low flexibility to allow tests of different cardiovascular devices and low precision when a reference model must be reproduced. In this paper a new bench is described. It combines the computer model of the environment in which the device will operate and the electro-hydraulic interfaces by which device and computer are connected. A models library provided with basic functions allows implementing many layouts of the bench, which in turn depend both on the device properties and the desired experiment. In case of an apical LVAD evaluation, the bench can reproduce right and left ventricles, pulmonary and systemic circulations, inlet and outlet LVAD cannulas. An interface forces the instantaneous calculated flow at the VAD input and feeds back the measured pressure to the computer; another interface works in a similar -but complementary- way at the VAD output. The paper focuses on the operating principle of the electro hydraulic interfaces which represent a relevant component of the bench, on the RT-Linux-based software architecture, on the models of the basic elements of the bench. A patent is under preparation. At the moment, only a portion of the bench has been developed. It consists of a piston-cylinder mechanism, which mimics the elastance-based mechanism of a natural ventricle, and a hydraulic circuit representing the arterial load according to a modified windkessel model and the venous return according to the Guyton's model. The pump is driven by a real-time simulation of the cardiovascular system. This preliminary layout allowed testing the piston-cylinder mechanism, its control, and the software. This electro-hydraulic interface has been used to reproduce a pulsatile pump working in different modes. The hybrid model approach can support the development of new cardiac assist devices from their computer model to their manufacture.