Stroke has become one of the leading causes of mortality worldwide and about 800 in every 100,000 people suffer from stroke each year. The occurrence of stroke is ranked third among the causes of acute death and first among the causes for neurological dysfunction. Currently, Neurological examinations followed by medical imaging with CT, MRI or Angiography are used to provide better identification of the location and the type of the stroke, however they are neither fast, cost-effective nor portable. Microwave technology has emerged to complement these modalities to diagnose stroke as it is sensitive to the differences between the distinct dielectric properties of the brain tissues and blood. This paper investigates the possibility of diagnosing the type of stroke using Finite Element Analysis (FEA). The object of interest is a simulated head phantom with stroke, created with its specifying material characteristics like electrical conductivity and relative permittivity. The phantom is then placed in an electromagnetic field generated by a dipole antenna radiating at 1 GHz. The FEM forward model solver computes the scattered electromagnetic field by finding the solution for the Maxwell's wave equation in the head volume. Subsequently the inverse scattering problem is solved using the Contrast Source Inversion (CSI) method to reconstruct the dielectric profile of the head phantom.