Speckle contrast optical tomography (SCOT) is a non-invasive imaging and real-time imaging technique for mapping cerebral blood flow (CBF). It has the potential for imaging at the bedside and has cost and signal-to-noise ratio (SNR) advantages over diffuse correlation spectroscopy (DCS), a popular optical technique for imaging CBF. To evaluate the performance of fiber-based SCOT using an anatomical model, we built a pipeline for simulating SCOT measurements and reconstructed images. We studied the reconstructed images under varied source-detector distances and system exposure times with a simulated perturbation of CBF and demonstrated that the cortical region of the brain can be measured. Simulated images showed reduced localization error when including larger source-detector distances. However, noise in an imaging optical system can prohibit the attempt. We also demonstrate that the system exposure time plays an important role in optimizing the SNR. These results will guide the design of future research instruments.