Kickingereder P, et al. - In this retrospective study, researchers assessed data from 455 subjects with brain tumors to establish a framework for fully automated quantitative analysis of MRI in neuro-oncology dependent on artificial neural networks (ANNs) to surmount the intrinsic limitations of manual assessment of tumor burden. In both longitudinal test datasets, the ANNs offered superior performance for accurate detection as well as segmentation of contrast-enhancing tumors and non-enhancing T2-signal abnormality volumes. They observed the need for 10 minutes of computation time (average per scan) for automated on-demand processing of MRI scans and quantitative tumor response assessment in the simulated clinical environment in the Heidelberg simulation dataset. They concluded that, in neuro-oncology, ANN allowed objective and automated assessment of tumor response at high throughput; this ultimately worked as a outline for the utilization of ANN in radiology to improve clinical decision making.