Deep learning-based analysis of macaque corneal sub-basal nerve fibers in confocal microscopy images

Jonathan D Oakley,Stuti L Misra,Jessica M Izzi,Megan E Mccarron,Rachel L Weinberg,Joseph L Mankowski,Daniel B Russakoff,Charles N Mcghee

doi:10.1186/s40662-020-00192-5

Jonathan D Oakley, Stuti L Misra + Show 6 more

Open Access

https://doi.org/10.1186/s40662-020-00192-5

Copy DOI

Abstract

BackgroundTo develop and validate a deep learning-based approach to the fully-automated analysis of macaque corneal sub-basal nerves using in vivo confocal microscopy (IVCM).MethodsIVCM was used to collect 108 images from 35 macaques. 58 of the images from 22 macaques were used to evaluate different deep convolutional neural network (CNN) architectures for the automatic analysis of sub-basal nerves relative to manual tracings. The remaining images were used to independently assess correlations and inter-observer performance relative to three readers.ResultsCorrelation scores using the coefficient of determination between readers and the best CNN averaged 0.80. For inter-observer comparison, inter-correlation coefficients (ICCs) between the three expert readers and the automated approach were 0.75, 0.85 and 0.92. The ICC between all four observers was 0.84, the same as the average between the CNN and individual readers.ConclusionsDeep learning-based segmentation of sub-basal nerves in IVCM images shows high to very high correlation to manual segmentations in macaque data and is indistinguishable across readers. As quantitative measurements of corneal sub-basal nerves are important biomarkers for disease screening and management, the reported work offers utility to a variety of research and clinical studies using IVCM.

Highlights

To develop and validate a deep learning-based approach to the fully-automated analysis of macaque corneal sub-basal nerves using in vivo confocal microscopy (IVCM)
As macaque models are used in a variety of diseases characterized by corneal sensory nerve fiber loss, we have developed and characterized a novel approach for automated analysis of nerve fibers, leveraging more current technologies in the world of computer vision and machine learning to process macaque IVCM images that are inherently of lower quality than human ICVM image [16,17,18,19,20]
This study reports on using similar deep learning-based architectures for the automated tracing of corneal nerve fibers in IVCM images of macaque corneas

Summary

Introduction

To develop and validate a deep learning-based approach to the fully-automated analysis of macaque corneal sub-basal nerves using in vivo confocal microscopy (IVCM). In vivo confocal microscopy (IVCM) of the cornea allows for non-invasive acquisition of two-dimensional images, enabling detailed corneal sensory nerve fiber assessment in both clinical and research settings. As noted by Dabbah [7], this lack of standardized assessment of corneal sub-basal nerve fiber density is a major limitation to wider adoption in clinical settings. Several different approaches have been used to automate the task of nerve fiber tracing in IVCM. This work has resulted in the freely available ACCMetrics tool, widely accepted as a standard in clinical IVCM image analysis [2, 7, 10,11,12,13]

Methods

Results

Conclusion