PocketNet: A Smaller Neural Network for Medical Image Analysis

Medical imaging deep learning models are often large and complex, requiring specialized hardware to train and evaluate these models. To address such issues, we propose the PocketNet paradigm to reduce the size of deep learning models by throttling the growth of the number of channels in convolutional neural networks. We demonstrate that, for a range of segmentation and classification tasks, PocketNet architectures produce results comparable to that of conventional neural networks while reducing the number of parameters by multiple orders of magnitude, using up to 90% less GPU memory, and speeding up training times by up to 40%, thereby allowing such models to be trained and deployed in resource-constrained settings.

[1]  Ziv Yaniv,et al.  Image Segmentation, Registration and Characterization in R with SimpleITK. , 2018, Journal of statistical software.

[2]  Kilian Q. Weinberger,et al.  Densely Connected Convolutional Networks , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[3]  Luis Ibáñez,et al.  The Design of SimpleITK , 2013, Front. Neuroinform..

[4]  Feng-Ping An,et al.  Medical image segmentation algorithm based on feedback mechanism convolutional neural network , 2019, Biomed. Signal Process. Control..

[5]  R. Meier,et al.  Fully automated brain resection cavity delineation for radiation target volume definition in glioblastoma patients using deep learning , 2020, Radiation oncology.

[6]  Sanjeev Arora,et al.  On the Optimization of Deep Networks: Implicit Acceleration by Overparameterization , 2018, ICML.

[7]  Alexander Wong,et al.  COVID-Net: a tailored deep convolutional neural network design for detection of COVID-19 cases from chest X-ray images , 2020, Scientific reports.

[8]  Jose Dolz,et al.  Multi-Scale Self-Guided Attention for Medical Image Segmentation , 2021, IEEE Journal of Biomedical and Health Informatics.

[9]  Nima Tajbakhsh,et al.  UNet++: A Nested U-Net Architecture for Medical Image Segmentation , 2018, DLMIA/ML-CDS@MICCAI.

[10]  Yann LeCun,et al.  Optimal Brain Damage , 1989, NIPS.

[11]  Jinchao Xu,et al.  MgNet: A unified framework of multigrid and convolutional neural network , 2019, Science China Mathematics.

[12]  Thomas Brox,et al.  3D U-Net: Learning Dense Volumetric Segmentation from Sparse Annotation , 2016, MICCAI.

[13]  Enhua Wu,et al.  Squeeze-and-Excitation Networks , 2017, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[14]  Nicholas Slevin,et al.  Evaluation of an automatic segmentation algorithm for definition of head and neck organs at risk , 2014, Radiation oncology.

[15]  Yousef Saad,et al.  Iterative methods for sparse linear systems , 2003 .

[16]  Jian Sun,et al.  Identity Mappings in Deep Residual Networks , 2016, ECCV.

[17]  Ziv Yaniv,et al.  SimpleITK Image-Analysis Notebooks: a Collaborative Environment for Education and Reproducible Research , 2017, Journal of Digital Imaging.

[18]  J. Zico Kolter,et al.  Overfitting in adversarially robust deep learning , 2020, ICML.

[19]  et al.,et al.  Identifying the Best Machine Learning Algorithms for Brain Tumor Segmentation, Progression Assessment, and Overall Survival Prediction in the BRATS Challenge , 2018, ArXiv.

[20]  Neeraj Sharma,et al.  Automated medical image segmentation techniques , 2010, Journal of medical physics.

[21]  Beatrice Riviere,et al.  Identification of kernels in a convolutional neural network: connections between the level set equation and deep learning for image segmentation , 2020, Medical Imaging: Image Processing.

[22]  Ahmed Ibrahim Alzahrani,et al.  Efficient 3D Deep Learning Model for Medical Image Semantic Segmentation , 2020 .

[23]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[24]  Yujie Li,et al.  NAS-Unet: Neural Architecture Search for Medical Image Segmentation , 2019, IEEE Access.

[25]  Yuanzhi Li,et al.  A Convergence Theory for Deep Learning via Over-Parameterization , 2018, ICML.

[26]  Joost van der Putten,et al.  Influence of decoder size for binary segmentation tasks in medical imaging , 2020, Medical Imaging: Image Processing.

[27]  Septimiu E. Salcudean,et al.  Reducing the Hausdorff Distance in Medical Image Segmentation With Convolutional Neural Networks , 2019, IEEE Transactions on Medical Imaging.

[28]  Qiegen Liu,et al.  X-Net: Brain Stroke Lesion Segmentation Based on Depthwise Separable Convolution and Long-range Dependencies , 2019, MICCAI.

[29]  Brian B. Avants,et al.  The Multimodal Brain Tumor Image Segmentation Benchmark (BRATS) , 2015, IEEE Transactions on Medical Imaging.

[30]  Sébastien Ourselin,et al.  Generalised Dice overlap as a deep learning loss function for highly unbalanced segmentations , 2017, DLMIA/ML-CDS@MICCAI.

[31]  François Chollet,et al.  Xception: Deep Learning with Depthwise Separable Convolutions , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[32]  R Cameron Craddock,et al.  The preprocessed connectomes project repository of manually corrected skull-stripped T1-weighted anatomical MRI data , 2016, bioRxiv.

[33]  Christos Davatzikos,et al.  Advancing The Cancer Genome Atlas glioma MRI collections with expert segmentation labels and radiomic features , 2017, Scientific Data.

[34]  Bo Chen,et al.  MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications , 2017, ArXiv.

[35]  Thomas Brox,et al.  U-Net: Convolutional Networks for Biomedical Image Segmentation , 2015, MICCAI.

[36]  Liqiang Zhang,et al.  3D Depthwise Convolution: Reducing Model Parameters in 3D Vision Tasks , 2018, Canadian Conference on AI.

[37]  Konstantinos Kamnitsas,et al.  Efficient multi‐scale 3D CNN with fully connected CRF for accurate brain lesion segmentation , 2016, Medical Image Anal..