5.5. Conclusion 5 81 Supplementary Figures Figure S1: Demonstration of the use of the simulator in conjunction with external software for receptive field prediction (Goebel et al., 2006). On the left, a 3D brain model with receptive field mapping (color indicates eccentricity in the visual field) is used to simulate the placement of four 8-by-8 Utah-like electrode arrays on the accessible parts of V1 in the left hemisphere. The inset shows a close up of the electrodes and the expected locations of the visual receptive fields (color indicates eccentricity). On the right we see the resulting simulation of stimulating these electrodes with a current amplitude of 70µA, after 167 milliseconds of continuous stimulation. To simulate imperfect knowledge of the electrode or phosphene locations, a small normally distributed noise (σ=0.03deg) was added to the determined receptive fields. Note that as the simulator is initialized with the phosphene locations in the visual field, any assumptions about feasible electrode locations, uncertainties and other sources of noise can be incorporated flexibly. Figure S2: Example simulation of irregular phosphene shapes. In our software, it is relatively straightforward to change the shape of individual phosphenes. The simulator includes the functionality of initializing with elongated phosphenes and the phosphene maps can also be manually adjusted to simulate arbitrary shapes. a) Schematic illustration of the modelled visual field location. b) The simulation output after initialization with 50 elongated phosphenes with random orientations. c) Simulation output with manually customized phosphene shapes.
RkJQdWJsaXNoZXIy MTk4NDMw