Format

Send to

Choose Destination
J Digit Imaging. 2019 Oct 25. doi: 10.1007/s10278-019-00282-4. [Epub ahead of print]

A Deep Convolutional Neural Network for Annotation of Magnetic Resonance Imaging Sequence Type.

Author information

1
Mathematical NeuroOncology Lab, Precision Neurotherapeutics Innovation Program, Mayo Clinic, 5777 East Mayo Blvd, Support Services Building Suite 2-700, Phoenix, AZ, 85054, USA. ranjbar.sara@mayo.edu.
2
Department of Neurosurgery, Mayo Clinic, Phoenix, AZ, USA. ranjbar.sara@mayo.edu.
3
Mathematical NeuroOncology Lab, Precision Neurotherapeutics Innovation Program, Mayo Clinic, 5777 East Mayo Blvd, Support Services Building Suite 2-700, Phoenix, AZ, 85054, USA.
4
Department of Neurosurgery, Mayo Clinic, Phoenix, AZ, USA.
5
Department of Biostatistics and Bioinformatics, H. Lee Moffitt Cancer Center and Research Institute, Tampa, FL, USA.
6
Department of Radiology, Mayo Clinic, Phoenix, AZ, USA.

Abstract

The explosion of medical imaging data along with the advent of big data analytics has launched an exciting era for clinical research. One factor affecting the ability to aggregate large medical image collections for research is the lack of infrastructure for automated data annotation. Among all imaging modalities, annotation of magnetic resonance (MR) images is particularly challenging due to the non-standard labeling of MR image types. In this work, we aimed to train a deep neural network to annotate MR image sequence type for scans of brain tumor patients. We focused on the four most common MR sequence types within neuroimaging: T1-weighted (T1W), T1-weighted post-gadolinium contrast (T1Gd), T2-weighted (T2W), and T2-weighted fluid-attenuated inversion recovery (FLAIR). Our repository contains images acquired using a variety of pulse sequences, sequence parameters, field strengths, and scanner manufacturers. Image selection was agnostic to patient demographics, diagnosis, and the presence of tumor in the imaging field of view. We used a total of 14,400 two-dimensional images, each visualizing a different part of the brain. Data was split into train, validation, and test sets (9600, 2400, and 2400 images, respectively) and sets consisted of equal-sized groups of image types. Overall, the model reached an accuracy of 99% on the test set. Our results showed excellent performance of deep learning techniques in predicting sequence types for brain tumor MR images. We conclude deep learning models can serve as tools to support clinical research and facilitate efficient database management.

KEYWORDS:

Artificial intelligence; Automated annotation; Deep learning; Image database; Magnetic resonance imaging; Sequence type

PMID:
31654174
DOI:
10.1007/s10278-019-00282-4

Supplemental Content

Full text links

Icon for Springer
Loading ...
Support Center