One of the most well-studied brain rhythms is the 11-15 Hz sleep “spindle” oscillation in non-REM sleep. In recent work, we studied sleep spindles in large-scale intracranial recordings from human clinical patents. By applying methods we have developed to track patterns in noisy multisite data to these recordings, we found that cortical spindles are spatiotemporally organized into circular wave-like patterns, travelling in a preferred direction. Because the propagation speed of the observed waves is consistent with the axonal conduction speed of the cortical white matter fibers, which are coated in myelin sheathes to increase conduction speed (3-5 m/s), we hypothesized these fibers are a potential substrate; however, how they create the observed spatiotemporal structures and their preferred direction of rotation was unclear. We now present results from a time-delay Kuramoto model demonstrating that the emergence of these spatiotemporal patterns is due to heterogeneous time delays in the system. The emergence of rotating waves with a preferred chirality is specific to the spindle oscillation frequency, suggesting that this is due to an interplay of the time-delays inherent in the system and the frequency of this sleep rhythm. We then introduce a complex-valued matrix formulation for the time-delay Kuramoto network dynamics, which provides insight into the network-level mechanism for these dynamics.