Deep Learning-Based Histopathologic Assessment of Kidney Tissue.

BACKGROUND The development of deep neural networks is facilitating more advanced digital analysis of histopathologic images. We trained a convolutional neural network for multiclass segmentation of digitized kidney tissue sections stained with periodic acid-Schiff (PAS). METHODS We trained the network using multiclass annotations from 40 whole-slide images of stained kidney transplant biopsies and applied it to four independent data sets. We assessed multiclass segmentation performance by calculating Dice coefficients for ten tissue classes on ten transplant biopsies from the Radboud University Medical Center in Nijmegen, The Netherlands, and on ten transplant biopsies from an external center for validation. We also fully segmented 15 nephrectomy samples and calculated the network's glomerular detection rates and compared network-based measures with visually scored histologic components (Banff classification) in 82 kidney transplant biopsies. RESULTS The weighted mean Dice coefficients of all classes were 0.80 and 0.84 in ten kidney transplant biopsies from the Radboud center and the external center, respectively. The best segmented class was "glomeruli" in both data sets (Dice coefficients, 0.95 and 0.94, respectively), followed by "tubuli combined" and "interstitium." The network detected 92.7% of all glomeruli in nephrectomy samples, with 10.4% false positives. In whole transplant biopsies, the mean intraclass correlation coefficient for glomerular counting performed by pathologists versus the network was 0.94. We found significant correlations between visually scored histologic components and network-based measures. CONCLUSIONS This study presents the first convolutional neural network for multiclass segmentation of PAS-stained nephrectomy samples and transplant biopsies. Our network may have utility for quantitative studies involving kidney histopathology across centers and provide opportunities for deep learning applications in routine diagnostics.

[1]  Rabi Yacoub,et al.  Unsupervised labeling of glomerular boundaries using Gabor filters and statistical testing in renal histology , 2017, Journal of medical imaging.

[2]  Jan U. Becker,et al.  A 2018 Reference Guide to the Banff Classification of Renal Allograft Pathology , 2018, Transplantation.

[3]  Bram van Ginneken,et al.  A survey on deep learning in medical image analysis , 2017, Medical Image Anal..

[4]  Michael Gadermayr,et al.  CNN cascades for segmenting sparse objects in gigapixel whole slide images , 2019, Comput. Medical Imaging Graph..

[5]  Michael Gadermayr,et al.  CNN Cascades for Segmenting Whole Slide Images of the Kidney , 2017, Comput. Medical Imaging Graph..

[6]  Wendy Moncur,et al.  The Accuracy and Reliability of Crowdsource Annotations of Digital Retinal Images , 2016, Translational vision science & technology.

[7]  P. Nickerson,et al.  Computerized image analysis of Sirius Red-stained renal allograft biopsies as a surrogate marker to predict long-term allograft function. , 2003, Journal of the American Society of Nephrology : JASN.

[8]  John E. Tomaszewski,et al.  Computational analysis of the structural progression of human glomeruli in diabetic nephropathy , 2018, Medical Imaging.

[9]  Gloria Bueno,et al.  Glomerulus Classification with Convolutional Neural Networks , 2017, MIUA.

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

[11]  John Tomaszewski,et al.  Digital pathology evaluation in the multicenter Nephrotic Syndrome Study Network (NEPTUNE). , 2013, Clinical journal of the American Society of Nephrology : CJASN.

[12]  Bryan F. Cox,et al.  Computer-assisted imaging algorithms facilitate histomorphometric quantification of kidney damage in rodent renal failure models , 2012, Journal of pathology informatics.

[13]  A. Rule,et al.  The Substantial Loss of Nephrons in Healthy Human Kidneys with Aging. , 2017, Journal of the American Society of Nephrology : JASN.

[14]  J. Olivo-Marin,et al.  Interstitial Fibrosis Evolution on Early Sequential Screening Renal Allograft Biopsies Using Quantitative Image Analysis , 2011, American journal of transplantation : official journal of the American Society of Transplantation and the American Society of Transplant Surgeons.

[15]  H. E. Hansen,et al.  The Banff 97 working classification of renal allograft pathology. , 1999, Kidney international.

[16]  Daniel A Beard,et al.  Region-Based Convolutional Neural Nets for Localization of Glomeruli in Trichrome-Stained Whole Kidney Sections. , 2018, Journal of the American Society of Nephrology : JASN.

[17]  Paul J Tadrous,et al.  On the concept of objectivity in digital image analysis in pathology , 2010, Pathology.

[18]  Michael Gadermayr,et al.  Do We Need Large Annotated Training Data for Detection Applications in Biomedical Imaging? A Case Study in Renal Glomeruli Detection , 2016, MLMI@MICCAI.

[19]  Nico Karssemeijer,et al.  Whole-Slide Mitosis Detection in H&E Breast Histology Using PHH3 as a Reference to Train Distilled Stain-Invariant Convolutional Networks , 2018, IEEE Transactions on Medical Imaging.

[20]  Neil Roberts,et al.  Medical Image Computing and Computer-Assisted Intervention -- MICCAI 2015 , 2015, Lecture Notes in Computer Science.

[21]  Yoshihiro Hirohashi,et al.  Segmental HOG: new descriptor for glomerulus detection in kidney microscopy image , 2015, BMC Bioinformatics.

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

[23]  T. Muthukumar,et al.  The Banff 2015 Kidney Meeting Report: Current Challenges in Rejection Classification and Prospects for Adopting Molecular Pathology , 2016, American journal of transplantation : official journal of the American Society of Transplantation and the American Society of Transplant Surgeons.

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