Back to Search
Start Over
A Perlin Noise-Based Augmentation Strategy for Deep Learning with Small Data Samples of HRCT Images
- Source :
- Scientific Reports, Scientific Reports, Vol 8, Iss 1, Pp 1-7 (2018)
- Publication Year :
- 2018
- Publisher :
- Springer Science and Business Media LLC, 2018.
-
Abstract
- Deep learning is now widely used as an efficient tool for medical image classification and segmentation. However, conventional machine learning techniques are still more accurate than deep learning when only a small dataset is available. In this study, we present a general data augmentation strategy using Perlin noise, applying it to pixel-by-pixel image classification and quantification of various kinds of image patterns of diffuse interstitial lung disease (DILD). Using retrospectively obtained high-resolution computed tomography (HRCT) images from 106 patients, 100 regions-of-interest (ROIs) for each of six classes of image patterns (normal, ground-glass opacity, reticular opacity, honeycombing, emphysema, and consolidation) were selected for deep learning classification by experienced thoracic radiologists. For extra-validation, the deep learning quantification of the six classification patterns was evaluated for 92 HRCT whole lung images for which hand-labeled segmentation masks created by two experienced radiologists were available. FusionNet, a convolutional neural network (CNN), was used for training, test, and extra-validation on classifications of DILD image patterns. The accuracy of FusionNet with data augmentation using Perlin noise (89.5%, 49.8%, and 55.0% for ROI-based classification and whole lung quantifications by two radiologists, respectively) was significantly higher than that with conventional data augmentation (82.1%, 45.7%, and 49.9%, respectively). This data augmentation strategy using Perlin noise could be widely applied to deep learning studies for image classification and segmentation, especially in cases with relatively small datasets.
- Subjects :
- Computer science
lcsh:Medicine
Computed tomography
Convolutional neural network
Article
030218 nuclear medicine & medical imaging
03 medical and health sciences
0302 clinical medicine
medicine
Segmentation
Honeycombing
lcsh:Science
Multidisciplinary
Small data
medicine.diagnostic_test
Contextual image classification
business.industry
Deep learning
lcsh:R
Interstitial lung disease
Pattern recognition
medicine.disease
030220 oncology & carcinogenesis
lcsh:Q
Artificial intelligence
Perlin noise
business
Subjects
Details
- ISSN :
- 20452322
- Volume :
- 8
- Database :
- OpenAIRE
- Journal :
- Scientific Reports
- Accession number :
- edsair.doi.dedup.....4491a840c4a801f573dbdd67e5c354e4