Plexus Convolutional Neural Network (PlexusNet): A novel neural network architecture for histologic image analysis

Different convolutional neural network (CNN) models have been tested for their application in histological image analyses. However, these models are prone to overfitting due to their large parameter capacity, requiring more data or valuable computational resources for model training. Given these limitations, we introduced a novel architecture (termed PlexusNet). We utilized 310 Hematoxylin and Eosin stained (HE it helped reduce the slide number to examine by 43.8% without consequences, although its parameter capacity was 200 times smaller than ResNet18. We found that the partitioning of the development set influences the model calibration for all models. However, with PlexusNet architecture, we could achieve comparable well-calibrated models trained on different partitions. In conclusion, PlexusNet represents a novel model architecture for histological image analysis that achieves classification performance comparable to other models while providing orders-of-magnitude parameter reduction.

[1]  Arkadiusz Gertych,et al.  An EM-based semi-supervised deep learning approach for semantic segmentation of histopathological images from radical prostatectomies , 2018, Comput. Medical Imaging Graph..

[2]  Sergey Ioffe,et al.  Rethinking the Inception Architecture for Computer Vision , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

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

[4]  Onno J de Boer,et al.  Deep learning for automatic Gleason pattern classification for grade group determination of prostate biopsies , 2019, Virchows Archiv.

[5]  A. Chandra,et al.  Challenges in Understanding Differences in Health Care Spending Between the United States and Other High-Income Countries. , 2018, Journal of the American Medical Association (JAMA).

[6]  Ewout W Steyerberg,et al.  Net benefit approaches to the evaluation of prediction models, molecular markers, and diagnostic tests , 2016, British Medical Journal.

[7]  B. van Ginneken,et al.  Deep learning as a tool for increased accuracy and efficiency of histopathological diagnosis , 2016, Scientific Reports.

[8]  Sergey Ioffe,et al.  Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift , 2015, ICML.

[9]  Yun Jiang,et al.  Breast cancer histopathological image classification using convolutional neural networks with small SE-ResNet module , 2019, PloS one.

[10]  Ashish K. Jha,et al.  Health Care Spending in the United States and Other High-Income Countries , 2018, JAMA.

[11]  T. Hermanns,et al.  Automated Gleason grading of prostate cancer tissue microarrays via deep learning , 2018, Scientific Reports.

[12]  Peter König,et al.  Data augmentation instead of explicit regularization , 2018, ArXiv.

[13]  Forrest N. Iandola,et al.  DenseNet: Implementing Efficient ConvNet Descriptor Pyramids , 2014, ArXiv.

[14]  Andrew J Vickers,et al.  The Brier score does not evaluate the clinical utility of diagnostic tests or prediction models , 2017, Diagnostic and Prognostic Research.

[15]  Karel J. Zuiderveld,et al.  Contrast Limited Adaptive Histogram Equalization , 1994, Graphics Gems.

[16]  Thomas Brox,et al.  U-Net: deep learning for cell counting, detection, and morphometry , 2018, Nature Methods.

[17]  Lei Chen,et al.  Automated Diagnosis of Lymphoma with Digital Pathology Images Using Deep Learning , 2018, Annals of clinical and laboratory science.

[18]  Sanjiv Kumar,et al.  Adaptive Methods for Nonconvex Optimization , 2018, NeurIPS.

[19]  Edward H. Adelson,et al.  The Design and Use of Steerable Filters , 1991, IEEE Trans. Pattern Anal. Mach. Intell..

[20]  E. Elkin,et al.  Decision Curve Analysis: A Novel Method for Evaluating Prediction Models , 2006, Medical decision making : an international journal of the Society for Medical Decision Making.

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

[22]  Thomas J. Fuchs,et al.  Clinical-grade computational pathology using weakly supervised deep learning on whole slide images , 2019, Nature Medicine.

[23]  Factors Contributing to Higher Health Care Spending in the United States Compared With Other High-Income Countries. , 2018, JAMA.

[24]  Jonathan Epstein,et al.  Grading of prostatic adenocarcinoma: current state and prognostic implications , 2016, Diagnostic Pathology.

[25]  Andrew H. Beck,et al.  Diagnostic Assessment of Deep Learning Algorithms for Detection of Lymph Node Metastases in Women With Breast Cancer , 2017, JAMA.

[26]  Bram van Ginneken,et al.  Epithelium segmentation using deep learning in H&E-stained prostate specimens with immunohistochemistry as reference standard , 2018, Scientific Reports.

[27]  Andrew Janowczyk,et al.  Deep learning for digital pathology image analysis: A comprehensive tutorial with selected use cases , 2016, Journal of pathology informatics.

[28]  Hadi Rezaeilouyeh,et al.  Microscopic medical image classification framework via deep learning and shearlet transform , 2016, Journal of medical imaging.

[29]  T. Hermanns,et al.  Author Correction: Automated Gleason grading of prostate cancer tissue microarrays via deep learning , 2019, Scientific Reports.

[30]  Andrew Zisserman,et al.  Very Deep Convolutional Networks for Large-Scale Image Recognition , 2014, ICLR.

[31]  Judith D. Cohn,et al.  Sparse coding of pathology slides compared to transfer learning with deep neural networks , 2018, BMC Bioinformatics.

[32]  Cyril Ferdynus,et al.  Decision curve analysis: a technical note. , 2018, Annals of translational medicine.