The MFE log-odds score, displayed in Figure 2D, were obtained by (5) s c o r e (M F E ) = l o g (P (a b s | p r e ) P (a b s | b g r ) ) The log of ratio of conditional probabilities of MFE of real and background precursors was computed for each of the bins, and were plotted.
Then, a similarity image between the reference background and training intensity images was computed using Equation 2 and the reference binary background image (RBBI) was computed: For each pixel i, j r i, j = 1 N ∑ t = 0 N - 1 SI RBI I, A t I i, j RBBI i, j = 0 static background if ( r ( i, j ) > 0.8 ) 1 moving background otherwise (8).
Mean intensities of the signal at this organ and background were computed and averaged for each image.
For each fluorescent image, the average pixel intensity within each circle was determined and a local background using mean pixel intensity was computed for each spot.
The average GO similarity score for each cluster was computed, and a total average similarity score was computed for each of the three background models.
After ROI delineation the spill-out region is identified and for each voxel inside the spill-out region the local background is computed (see below).
It was desirable that all arrays used for expression profiling have similar background scores, so background scores were computed for all 5,700 arrays, and those with scores below percentile 7.5 and above percentile 92.5 were eliminated.
A background model is computed for each camera view, using a mixture of Gaussians to estimate the background color per pixel.
The centroids of the particles are then located using the distance transformation, where the distance to the background region is computed for each foreground (particle) pixel.
To compensate for this phenomenon, a background distribution is computed for each described pattern based on a permuted set of protein structures.
