More articles from FUNCTIONAL

This retrospective study included preoperative MR imaging of 288 pediatric patients with pediatric posterior fossa tumors, including medulloblastoma (n=111), ependymoma (n=70), and pilocytic astrocytoma (n=107). Radiomics features were extracted from T2-weighted images, contrast-enhanced T1-weighted images, and ADC maps. Models generated by standard manual optimization by a machine learning expert were compared with automatic machine learning via the Tree-Based Pipeline Optimization Tool for performance evaluation. The authors conclude that automatic machine learning based on routine MR imaging classified pediatric posterior fossa tumors with high accuracy compared with manual expert pipeline optimization and qualitative expert MR imaging review.

Using the Alzheimer's Disease Neuroimaging Initiative dataset, the authors identified 2582 18F-florbetapir PET scans, which were separated into positive and negative cases by using a standardized uptake value ratio threshold of 1.1. They trained convolutional neural networks to predict standardized uptake value ratio and classify amyloid status. The best performance was seen for ResNet-50 by using regression before classification, 3 input PET slices, and pretraining, with a standardized uptake value ratio root-mean-squared error of 0.054, corresponding to 95.1% correct amyloid status prediction. The best trained network was more accurate than humans (96% versus a mean of 88%, respectively). They conclude that deep learning algorithms can estimate standardized uptake value ratio and use this to classify 18F-florbetapir PET scans and have promise to automate this laborious calculation.