Ultrasound tomography is an imaging technique for the breast that can create maps of the sound speed and acoustic attenuation simultaneously. By use of wave-based methods, images of speed of sound are of high quality, but the attenuation images are of inferior quality. The Spectral Log Difference (SLD) is a technique based on the ultrasonic backscatter that can provide estimates of the attenuation coefficient slope (ACS). However, SLD requires a large amount of data to provide a low variance of the estimate. This estimate variance can be reduced through compounding, which decreases spatial resolution of SLD images. Due to the nature of tomography, the tradeoff between estimate variance of SLD and spatial resolution can be reduced by using full angular spatial compounding. Using the data from a breast tomography scanner, the results demonstrated that the SLD technique with full angular spatial compounding could generate an accurate attenuation map with low bias and variance. As a result, SLD with full spatial angular compounding can provide improved attenuation images for breast diagnostics.