Combined Transfer Learning and Test-Time Augmentation Improves Convolutional Neural Network-Based Semantic Segmentation of Prostate Cancer from Multi-Parametric MR Images

•Test-time augmentation with transfer learning significantly improved CNN performance.•Results were statistically significant even in a small biomedical image data set.•ADC was found to be the most important MRI contrast, followed by Ktrans and T2w. Multiparametric MRI (mp-MRI) is a widely used tool...

Full description

Saved in:
Bibliographic Details
Published in:Computer methods and programs in biomedicine 2021-10, Vol.210, p.106375-106375, Article 106375
Main Authors: Hoar, David, Lee, Peter Q., Guida, Alessandro, Patterson, Steven, Bowen, Chris V., Merrimen, Jennifer, Wang, Cheng, Rendon, Ricardo, Beyea, Steven D., Clarke, Sharon E.
Format: Article
Language:English
Subjects:
Computer aided diagnosis
Convolutional neural network
Machine learning
MRI
Prostate cancer
Segmentation
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:•Test-time augmentation with transfer learning significantly improved CNN performance.•Results were statistically significant even in a small biomedical image data set.•ADC was found to be the most important MRI contrast, followed by Ktrans and T2w. Multiparametric MRI (mp-MRI) is a widely used tool for diagnosing and staging prostate cancer. The purpose of this study was to evaluate whether transfer learning, unsupervised pre-training and test-time augmentation significantly improved the performance of a convolutional neural network (CNN) for pixel-by-pixel prediction of cancer vs. non-cancer using mp-MRI datasets. 154 subjects undergoing mp-MRI were prospectively recruited, 16 of whom subsequently underwent radical prostatectomy. Logistic regression, random forest and CNN models were trained on mp-MRI data using histopathology as the gold standard. Transfer learning, unsupervised pre-training and test-time augmentation were used to boost CNN performance. Models were evaluated using Dice score and area under the receiver operating curve (AUROC) with leave-one-subject-out cross validation. Permutation feature importance testing was performed to evaluate the relative value of each MR contrast to CNN model performance. Statistical significance (p
ISSN:0169-2607
1872-7565
DOI:10.1016/j.cmpb.2021.106375