The authors recently spotted two errors in this article. In the Methods section, “Image processing” subheading, the sentence “Data sets were rejected if head displacement was greater than 5 mm in any direction” should read “Data sets were rejected if head displacement was greater than 2 mm in any direction”. In Fig. 4, an error occurred when transcribing the data point in the extreme right in the top right scatterplot (“Contralesional M1”). The statistical significance of the correlation is unchanged (Spearman's rho=+0.618, p=0.0088). The overall interpretation of the study is therefore unchanged. The new figure is printed here for the reader's convenience.