Fig. 2: MSI-SEER performance for colorectal cancer. | npj Digital Medicine

Fig. 2: MSI-SEER performance for colorectal cancer.

From: Deep Gaussian process with uncertainty estimation for microsatellite instability and immunotherapy response prediction from histology

Fig. 2: MSI-SEER performance for colorectal cancer.

The 3-fold cross-validation (CV) performance is evaluated for the training data, while the inter-cohort performance is evaluated for the validation datasets. Area under the ROC curve (AUC) values are shown. MSI-SEER, ResNet, EfficientNet, and ShuffleNet were trained using TCGA-CRC, Yonsei-1, and the combined data of TCGA-CRC and Yonsei-1. The 3-fold CV performance of MSIDETECT and MSI-SEER on TCGA-CRC was not evaluated because this dataset was already included in the training data for MSIDETECT. (*) denotes the dataset used for training and the remaining datasets were used for validation.

Back to article page