Optical coherence tomography (OCT) has many uses in medicine and engineering biology. It is a non-invasive technique for looking at the layered structures of tissues, such as skin, retina, teeth, and heart. Scanning of the retinal image reveals defects in the underlying layers. OCT performs in situ high resolution cross sectional imaging on a micron scale in real time. An application of OCT is in the imaging of diseases such as central serous retinopathy (CSR), which is the result of fluid accumulation under the macula. In the acquired OCT image, an inherent characteristic of coherent imaging is the presence of speckle noise. Reduction of speckle noise is one of the most important considerations for increasing the quality of coherent images. To enhance the quality, different filtering techniques are applied to analyze spectral noise in a CSR-OCT image. To analyze the effectiveness of the applied filtering techniques, various statistical parameters are applied and studied, such as the mean square error, peak signal to noise ratio, normalized crosscorrelation, and normalized absolute error. In this study, the resultant statistical information from the applied statistical parameters is subjected to multidimensional analysis to determine the best filtering technique to reduce speckle noise in CSR-OCT images.