Skip to main content
. 2019 Jun;74:12–24. doi: 10.1016/j.compmedimag.2019.02.006

Fig. 6.

Fig. 6

Pattern of the classification performance of the best models (i.e. for which the accuracy was above 80%) per stroke subtype (i.e. no stroke, large cortical, small cortical or lacunar) (left) and per stroke occurrence (i.e. had stroke or not) (right).