Table 1.
Baseline demographic and clinical characteristics of study sample.
Fig 1.
Flowchart of patient selection.
Fig 1 presents the distribution of patients after imputing the missing values. The labels on the arrows indicate the number of years of follow-up required in the training dataset and which filters were applied. The first box indicates the total number of patients assessed. Note that P (and %P) refers to the number (and percentage) of patients who meet “progression” criteria within the different subgroups.
Table 2.
Predictors of disease classification.
Table 3.
Predictive accuracy for different groups using 12 month visit without MRI based on ten-fold cross-validation.
Table 4.
Predictive accuracy for different groups using 12 month visit with MRI based on ten-fold cross-validation.
Table 5.
Predictive accuracy for different groups using 24 month visit without MRI based on ten-fold cross-validation.
Table 6.
Predictive accuracy for different groups using 24 month visit data with MRI based on ten-fold cross-validation.
Table 7.
Comparison of predictive accuracy for Group 2 (G2) using 12 month and 24 month clinical and MRI data based on ten-fold cross-validation.