One of the most common effects of a stroke is aphasia or the loss of the ability to comprehend and/or produce language. This accounts for much of the morbidity related to strokes.
In a study published by a group in Italy there is some convincing evidence that using tDCS can improve a naming task by 33% +/- 13%. This means that patients could demonstrate twenty to almost fifty percent improvement in naming. When you translate this into real-world recovery from stroke, the results are impressive.
As a tool for rehabilitation, and I am a physiatrist by trade (a rehabilitation doctor that directs the rehab of patients with diagnoses that include stroke), the clinical implications are significant.