A comparative performance evaluation of independent component analysis in medical image denoising

Medical images are often corrupted by noise arising in image acquisition process. Accurate diagnosis of the disease requires that medical images be sharp, clear and free of noise. Thus, image denoising is one of the fundamental tasks required by medical image analysis. There exist several denoising techniques for medical images like Median, Wavelet, Wiener, Average and Independent component analysis (ICA) filters. The independent component analysis is a statistical and computational technique for revealing hidden factors that underlie sets of random variables, measurements, or signals. In this paper, ICA has been used to separate out noise from the image to provide important diagnostic information to the physician and its usefulness is demonstrated by comparing its performance with other noise filtering methods. The performance of the ICA and other denoising techniques is evaluated using the metrics like Peak Signal-to-Noise Ratio (PSNR), Mean Absolute Error (MAE) and Mean Structural Similarity Index (MSSIM). The ICA based noise filtering technique gives 25.8245 dB of PSNR, 0.7312 of MAE and 0.9120 of SSIM. The experimental results and the performance comparisons show that ICA proves to be the effective method in eliminating noise from the medical image.