Directional Folded Antenna for Brain Stroke Detection Based on Classification Algorithm

To meet the demands on wideband unidirectional antennas in distinguishing ICH and IS stroke based on classification algorithm, a novel folded antenna is presented. The design of the antenna starts with a planar diplo antenna with bridge-shape ground. Then some slots are added to the ground to expand the bandwidth. Next, a folding technology is applied to the monopole to enhance its directivity while the overall size is reduced markedly at the same time. At last, a reflector and a pair of top-facing p-shape parasitic patches are applied to the antenna to increase the gain and decrease the voltage standing wave ratio (VSWR) at 1.7GHz. The final designed antenna has a dimensions of 0.293 lambda x0.107 lambda x0.053. (lambda is the wavelength at the lowest operating frequency). It has a measured reflection coefficient less than -10dB fractional bandwidth of 76% (1.1-2.45 GHz) and a peak gain of 4.5 dBi. In order to validate the effectiveness of the proposed antenna in the application, the proposed folded antenna is simulated with the human brain model and the result is classified by PCA and LDA classification algorithms. Finally, the accuracy of the final classification result is 100%, which indicates that the proposed folded antenna can be well applied to the brain stroke detection system.