Data Augmentation and Transfer Learning to Improve Generalizability of an Automated Prostate Segmentation Model.

OBJECTIVE. Deep learning applications in radiology often suffer from overfitting, limiting generalization to external centers. The objective of this study was to develop a high-quality prostate segmentation model capable of maintaining a high degree of performance across multiple independent datasets using transfer learning and data augmentation. MATERIALS AND METHODS. A retrospective cohort of 648 patients who underwent prostate MRI between February 2015 and November 2018 at a single center was used for training and validation. A deep learning approach combining 2D and 3D architecture was used for training, which incorporated transfer learning. A data augmentation strategy was used that was specific to the deformations, intensity, and alterations in image quality seen on radiology images. Five independent datasets, four of which were from outside centers, were used for testing, which was conducted with and without fine-tuning of the original model. The Dice similarity coefficient was used to evaluate model performance. RESULTS. When prostate segmentation models utilizing transfer learning were applied to the internal validation cohort, the mean Dice similarity coefficient was 93.1 for whole prostate and 89.0 for transition zone segmentations. When the models were applied to multiple test set cohorts, the improvement in performance achieved using data augmentation alone was 2.2% for the whole prostate models and 3.0% for the transition zone segmentation models. However, the best test-set results were obtained with models fine-tuned on test center data with mean Dice similarity coefficients of 91.5 for whole prostate segmentation and 89.7 for transition zone segmentation. CONCLUSION. Transfer learning allowed for the development of a high-performing prostate segmentation model, and data augmentation and fine-tuning approaches improved performance of a prostate segmentation model when applied to datasets from external centers.

[1]  Dong Yang,et al.  When Unseen Domain Generalization is Unnecessary? Rethinking Data Augmentation , 2019, ArXiv.

[2]  Clayton P Smith,et al.  A Grading System for the Assessment of Risk of Extraprostatic Extension of Prostate Cancer at Multiparametric MRI. , 2019, Radiology.

[3]  Bo Du,et al.  Exploiting Interslice Correlation for MRI Prostate Image Segmentation, from Recursive Neural Networks Aspect , 2018, Complex..

[4]  Daguang Xu,et al.  3D Anisotropic Hybrid Network: Transferring Convolutional Features from 2D Images to 3D Anisotropic Volumes , 2017, MICCAI.

[5]  Luis Perez,et al.  The Effectiveness of Data Augmentation in Image Classification using Deep Learning , 2017, ArXiv.

[6]  Alexander Wong,et al.  Fully automated segmentation of prostate whole gland and transition zone in diffusion-weighted MRI using convolutional neural networks , 2017, Journal of medical imaging.

[7]  Nathan Lay,et al.  Automatic magnetic resonance prostate segmentation by deep learning with holistically nested networks , 2017, Journal of medical imaging.

[8]  Bo Du,et al.  Deeply-supervised CNN for prostate segmentation , 2017, 2017 International Joint Conference on Neural Networks (IJCNN).

[9]  Daniel L. Rubin,et al.  Differential Data Augmentation Techniques for Medical Imaging Classification Tasks , 2017, AMIA.

[10]  Taghi M. Khoshgoftaar,et al.  A survey of transfer learning , 2016, Journal of Big Data.

[11]  Jian Sun,et al.  Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[12]  Allan Hanbury,et al.  Metrics for evaluating 3D medical image segmentation: analysis, selection, and tool , 2015, BMC Medical Imaging.

[13]  Jinkoo Kim,et al.  Contouring variability of human- and deformable-generated contours in radiotherapy for prostate cancer , 2015, Physics in medicine and biology.

[14]  Florian Jung,et al.  Evaluation of prostate segmentation algorithms for MRI: The PROMISE12 challenge , 2014, Medical Image Anal..

[15]  John Hudson,et al.  Into the valley of death: research to innovation. , 2013, Drug discovery today.

[16]  Tufve Nyholm,et al.  Variability in prostate and seminal vesicle delineations defined on magnetic resonance images, a multi-observer, -center and -sequence study , 2013, Radiation oncology.

[17]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[18]  Chinyere N. Onyebuchi,et al.  The role of preoperative endorectal magnetic resonance imaging in the decision regarding whether to preserve or resect neurovascular bundles during radical retropubic prostatectomy , 2004, Cancer.