Computational study on spatially distributed sequential stimulation (SDSS)

Silviu Agotici 1, 2, 3, 5, 6 ; Yoo, Paul 2, 3, 4, 6, 7 ; Masani, Kei 1, 2, 3, 5

1.   Toronto Rehabilitation Institute (TRI);    2.   University Health Network (UHN);    3.   Institute of Biomaterials & Biomedical Engineering;    4.   Edward S. Rogers Sr. Department of Electrical & Computer Engineering;    5.   Rehabilitation Engineering Laboratory;    6.   Neural Engineering Laboratory;    7.   Collaborative Program In Neuroscience (CPIN) 

Introduction and Aim: Functional electrical stimulation (FES) is used to restore limb or body movements in individuals with neurological impairments such as spinal cord injury or stroke. Currently, the clinical utility of FES-evoked muscle contractions during therapy is limited by the rapid onset of fatigue. To date, several methods have been proposed to circumvent FES-induced muscle fatigue, none have proven suitable for clinical applications. We have developed a novel strategy called spatially distributed sequential stimulation (SDSS), which minimizes fatigue by asynchronously stimulating different volumes of a targeted muscle. Preliminary studies in human subjects have shown promising resistance to stimulation-evoked muscle fatigue, but further improvements are needed. In this study, we aim to improve the SDSS approach by better understanding the mechanism by which SDSS activates different groups of muscle fibers.

Methods: 26 MRI images from a healthy individual were segmented and lofted together in Autodesk Inventor to create a 3D model of the lower leg. The MRI slices are all 8 mm apart and so the model depicts a 20 cm segment of the lower leg, located between 9.6 cm to 29.6 cm above the lateral malleolus. LiveLink for Inventor was used to import the CAD model into COMSOL Multiphysics where two electrodes were added to the surface of the model. The electrodes, modeled as cylinders of 25 mm diameter and 2mm thickness were placed 14 cm apart on the anterior side of the lower leg, where they would be closest to the TA muscle. The 3D model included 9 major muscles, the tibia and fibula, connective tissue, and two electrodes, the anode and the cathode. The anode was modeled as a current source having a normal current density of 20 A/m2 and the cathode was set as the ground.

Results: The finite element model (FEM) was built using extra fine free tetrahedral elements resulting in a mesh with 3,623,695 domain elements. A stationary study was run and the electric potential (V) and current density throughout the 3D model was solved for. To date, we have used FEA to observe how current flow and voltage distribution varies throughout the TA muscle relative to the normal current density at the anode, and relative to the placement of the anode and cathode electrodes on the surface of the skin.


Conclusion: Seeing how much electrical current flows through various regions of the TA muscle can give valuable insight into where the optimal placement of electrodes is. Since we would like to have electrical current focused on specific areas in the TA muscle, we are investigating what the optimal placement and size of electrodes will result in the most focused current density. So far the FEM model has given us good qualitative results allowing for some good initial inferences to be made.