Figure 3 | Scientific Reports

Figure 3

From: Effect of head motion-induced artefacts on the reliability of deep learning-based whole-brain segmentation

Figure 3

Boxplots showing the distributions of the dice similarity coefficient (DSC), intersection over union (IoU), volumetric difference (DF), and Hausdorff distance (HD) for FreeSurfer and deep learning-based segmentation methods. Evaluation metrics were calculated to compare the segmentation masks generated for good quality images acquired under rest to the masks generated for good (\({M}_{HM1}\)—\({M}_{HM1}\)), medium (\({M}_{HM1}\)—\({M}_{HM2}\)), and bad quality images (\({M}_{HM1}\)—\({M}_{HM3}\)) acquired under active head motion.

Back to article page