Image Level Training and Prediction: Intracranial Hemorrhage Identification in 3D Non-Contrast CT

Current hardware restrictions pose limitations on the use of convolutional neural networks for medical image analysis. There is a large trade-off between network architecture and input image size. For this reason, identification and classification tasks are commonly approached with patch or region-b...

Full description

Bibliographic Details
Main Authors: Ajay Patel, Sil. C. van de Leemput, Mathias Prokop, Bram Van Ginneken, Rashindra Manniesing
Format: Article
Language:English
Published: IEEE 2019-01-01
Series:IEEE Access
Subjects:
CAD
CNN
CT
Online Access:https://ieeexplore.ieee.org/document/8758429/