Study on the Effect of Noise in Microwave Tomography

In this paper, the capability of Exact Algorithm for operating in strongly noisy environment has been studied for reconstruction of images of biological target in Microwave Tomography. Other commonly used imaging system e.g. Ultrasound assisted tomography and Magnetic Resonance Imaging system require sophisticated Wavelet domain method of filtering technique for suppression of noise to improve the quality of image for medical purpose. Here, the effect of noise on the reconstruction was investigated by deliberately adding Additive White Gaussian Noise theoretically in the complex received electric field of the Microwave Imaging system for signal to noise ratio of 10dB to 60dB. For diseased model, percentage accuracy have changed from 90% - 100% for same variation of S/N ratio. Least accuracy is obtained in diseased organ kidney where complex permittivity value deviates from its normal value by 15% due to disease. This also indicates the presence of disease at that particular organ. From this study, it can be concluded that the accuracy of reconstructed image is found to be independent of noise.