The efficacy and cognitive outcomes of electro-convulsive therapy (ECT) on psychiatric disorders have been shown to depend on variations in treatment technique. In order to investigate this, a high resolution finite element human head model was generated from MRI scans and implemented with tissue heterogeneity and an excitable ionic neural formulations in the brain. The model was used to compare the effects of altered ECT stimulus amplitude and pulse width on the spatial extent of directly activated brain regions. The results showed that decreases in both amplitude and pulse width could effectively lead to reductions in the size of activated brain regions.