The study about filter patterns and the effect of each filter are performed by using Jaszczak(ellipse) and Hoffman phantoms. Six filters with various parameters are used for SPECT reconstruction Two terns, contrast and root mean square uncertainty(Xres) are calculated by using 9 pixels ROI for quantitative study and to obtain the selected parameter for each filter with Ramp as an ECT filter The contrast is proportional to root mean square uncertainty Moreover, the qualitative study is achieved by scoring from 21 observers opinions on the most desirable image. From the two studies, the compromisation between contrast and noise (%rms) is determined for optimal filter in SPECT brain reconstruction Wiener filter at FWHN = 3 and c = 0.10 preserved contrast while increased not much noise , therefore, this parameter of Wiener filter is the optimal filter for SPECT brain imaging particular for Toshiba system model GCA-901A.