Comparative Analysis of Image Enhancement Techniques for Ultrasonic Ovarian Cyst Images

The diagnosis of an ovarian cyst from an ultrasonic image is challenging due to inherent noises caused due to movements. The speckle noises present in the ultrasound degrades the image. Ultrasound imaging and analyzing the images for any cyst is the current method for the diagnosis of ovarian cysts. Thus an effective image preprocessing to reduce the noises and enhance the image is necessary for an accurate diagnosis of ovarian cysts. This work studies the different image filtering mechanisms namely median filter, fourier filter, butterworth filter, and daubechies wavelet filter to filter noises from ultrasonic ovarian images and analyzes the effectiveness of them for proper visualization of ovarian cysts. The different filter evaluation is done by measuring different performance of the filter metrics for image quality with respect to Mean Square Error(MSE), Signal-to-noise ratio(SNR), and Peak Signal to Noise ratio(PSNR). The wavelet filter showed a least value of MSE and SNR and PSNR are more compared to the other filters used. The wavelet filter is best among other filter in respect of MSE, SNR and PSNR.