This work presents an improved approach to characterize human head tissue by the microwave imaging technique. Microwave imaging is a non-invasive and non-ionizing imaging technique that employs low-power microwave signals. The detector is a highly directional Antipodal Vivaldi Antenna with a trapezoidal parasitic patch at the aperture and operates at a resonance frequency of 3.2 GHz. The interaction of the electromagnetic wave with human head tissue is analyzed by modelling a heterogeneous head mimicking phantom whose dielectric properties are like human head tissue. The tumor tissue is detected by analyzing reflected signals from an antenna. Tumor tissue produces strong reflections compared to surrounding healthy tissues because of changes in permittivity and conductivity. The size of the tumor is measured by the Resonant Frequency Shifting technique and the depth of the tumor is precisely detected by the Ground Penetrating Radar algorithm.