CONICET | Buscador de Institutos y Recursos Humanos

One of the key challenges in dense array transcranial electrical neuromodulation is to find a current injection pattern that maximizes the current density at a cortical region of interest and minimizes it in the rest of the head, which is mathematically modelled as an optimization problem. These optimal patterns can be found with help of the Least Squares (LS) or Linearly Constrained Minimum Variance (LCMV) algorithms. However, these methods return currents for all available electrodes in a dense array head harness. Therefore, each electrode requires to be driven by an independent current source, which is not always practical and at least expensive. The aim of this study is to analyze several optimization methods that comply with hardware and safety constraints. The reciprocity principle is instrumental in finding optimal injection patterns guided by the solution of the forward problem in Electroencephalography and taking into account the mentioned constrains. We have compared the performance of the reciprocity family methods in terms of bias, focality, intensity, and directionality using the LS and LCMV solutions as the baseline. Four targets on the cortex differing by depth and orientation were defined in a detailed seven-tissue MRI/CT based Finite Element head model. The optimal current injection patterns were computed using the LS, LCMV and three reciprocity derivative methods assuming 128 and 256 electrode nets, one current injection source and no fewer than 10 electrodes engaged as sources or as sinks from the safety considerations. All reciprocity algorithms showed good performance comparable to the LCMV and LS solutions. The best ?ad hoc? reciprocity version has demonstrated the focality similar to the LCMV and LS solutions, and higher intensity on the target in the desired direction. Comparing the 128 and 256 electrode cases we found that use of the larger electrode number improves the focality and intensity parameters.

enviar mensaje