Erratum

This is a correction to Yogananda CG, Shah BR, Nalawade SS, et al. MRI-based deep-learning method for determining glioma MGMT promoter methylation status. AJNR Am J Neuroradiol 2021;42:845–52 [10.3174/ajnr.A7029] [33664111]

There was an error in the Python code for the 3-fold cross-validation procedure. This resulted in the use of the training cases instead of the set-aside test cases for the testing procedure for molecular marker accuracy. This caused our reported accuracies from the TCIA/TCGA data set to be artificially inflated. The corrected accuracies for the Table (computed using nnU-Net¹), along with the updated receiver operating characteristic (ROC) curve for Fig 3 are provided here. The updated accuracies do not outperform other reported methods for MGMT molecular marker prediction using MR imaging.

Cross-validation results

Fold Description	MGMT-Net
	% Accuracy	AUC	Dice Score
Fold no.
Fold 1	59.75	0.4966	0.7906
Fold 2	73.49	0.6588	0.7725
Fold 3	64.63	0.5854	0.7874
Average	65.95 (SD, 0.06)	0.5802 (SD, 0.081)	0.7835 (SD, 0.009)

FIG 3.

ROC analysis for MGMT-net. Separate curves are plotted for each cross-validation fold along with corresponding area under the curve (AUC) values.