Fig. 1From: Estimation of brain amyloid accumulation using deep learning in clinical [11C]PiB PET imagingOverview of k-fold cross-validation training and ensemble predictions. The training/validation set consisted of 75% of the PET/CT scans. Hold-out test set was made up by the remaining 25% PET/CT scans (n = 300), all the PET/MRI scans (n = 100), and a set of challenging patients with significant co-morbidity scanned on a PET/CT (n = 30) or a PET/MRI (n = 7)Back to article page