Microwave imaging for stroke detection: validation on head-mimicking phantom

This paper provides initial results on the efficacy of Huygens Principle (HP) microwave imaging for haemorrhagic stroke detection. This is done using both simulations and measurements in an anechoic chamber. Microstrip antennas operating between 1 and 2 GHz have been designed, constructed and used for imaging a human head model in Computer Simulation Technology (CST) software. A 3D model consisting of human head tissues of Ella is employed in the simulation. An emulated haemorrhagic stroke with the dielectric properties equivalent to the blood has been inserted in Ella. Moreover, a 3-layered head-mimicking phantom containing an inclusion has been constructed. Frequency-domain measurements have been performed in an anechoic chamber using a Vector Network Analyser arrangement to obtain the transfer function (S21) between two antennas. Both simulations and measurements show that the HP based technique may be used for haemorrhagic stroke detection. Among linear scattering techniques, the HP based technique allows to detect dielectric inhomogeneities in the frequency domain. HP can also be used if the antennas and phantom are in free space, i.e. no coupling liquid is required. Detection of the haemorrhagic stroke has been achieved after removing the artefacts. Artefact removal is an essential step of any microwave imaging system and current artefact removal approaches have been shown to be ineffective in the specific scenario of brain imaging. However, one of this paper’s novel contributions is the proposal of an artefact removal algorithm based on a subtraction between S21 obtained using measurements, which achieves improved performance while having a much lower computational complexity.