Super‐resolution musculoskeletalMRI using deep learning
| Autor: | Akshay S. Chaudhari, Brian A. Hargreaves, Garry E. Gold, Jeffrey P. Wood, Zhongnan Fang, Eric K. Gibbons, Feliks Kogan, Kathryn J. Stevens, Jin Hyung Lee |
|---|---|
| Rok vydání: | 2018 |
| Předmět: |
Computer science
Image quality Pilot Projects Signal-To-Noise Ratio Residual Convolutional neural network Article 030218 nuclear medicine & medical imaging Upsampling 03 medical and health sciences Deep Learning Imaging Three-Dimensional 0302 clinical medicine Humans Knee Radiology Nuclear Medicine and imaging Root-mean-square deviation Phantoms Imaging business.industry Deep learning Pattern recognition Osteoarthritis Knee Magnetic Resonance Imaging Tricubic interpolation Artificial intelligence business Cartilage Diseases Algorithms 030217 neurology & neurosurgery Interpolation |
| Zdroj: | Magnetic Resonance in Medicine. 80:2139-2154 |
| ISSN: | 1522-2594 0740-3194 |
| DOI: | 10.1002/mrm.27178 |
| Popis: | PURPOSE To develop a super-resolution technique using convolutional neural networks for generating thin-slice knee MR images from thicker input slices, and compare this method with alternative through-plane interpolation methods. METHODS We implemented a 3D convolutional neural network entitled DeepResolve to learn residual-based transformations between high-resolution thin-slice images and lower-resolution thick-slice images at the same center locations. DeepResolve was trained using 124 double echo in steady-state (DESS) data sets with 0.7-mm slice thickness and tested on 17 patients. Ground-truth images were compared with DeepResolve, clinically used tricubic interpolation, and Fourier interpolation methods, along with state-of-the-art single-image sparse-coding super-resolution. Comparisons were performed using structural similarity, peak SNR, and RMS error image quality metrics for a multitude of thin-slice downsampling factors. Two musculoskeletal radiologists ranked the 3 data sets and reviewed the diagnostic quality of the DeepResolve, tricubic interpolation, and ground-truth images for sharpness, contrast, artifacts, SNR, and overall diagnostic quality. Mann-Whitney U tests evaluated differences among the quantitative image metrics, reader scores, and rankings. Cohen's Kappa (κ) evaluated interreader reliability. RESULTS DeepResolve had significantly better structural similarity, peak SNR, and RMS error than tricubic interpolation, Fourier interpolation, and sparse-coding super-resolution for all downsampling factors (p |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Plný text