Brain-Computer Interfaces convey a great potential in the field of stroke rehabilitation, where the continuous monitoring of the execution of mental tasks could support the positive effects of the therapy by reinforcing specific mental patterns. We propose the use of EEG Microstates as the building blocks of a novel BCI for Operators, and we show that preliminary results on a stroke patient seem to confirm that topographic information can provide hints on salient activity patterns on the affected hemisphere not recognizable within a simple discriminant framework. Furthermore, the technique could easily keep track of the evolution of these maps over different rehabilitation sessions, thus providing a strong hint of plasticity phenomena and cortical reorganization.