Automatic Evaluation of Corneal Nerve Tortuosity in Images from In Vivo Confocal Microscopy

Abstract

PURPOSE. An algorithm and a computer program for the automatic grading of corneal nerve tortuosity are proposed and evaluated. METHODS. Thirty images of the corneal subbasal nerve plexus with different grades of tortuosity were acquired with a scanning laser confocal microscope in normal and pathologic subjects. Nerves were automatically traced with an algorithm previously developed, and a tortuosity measure was computed with the proposed method, based on the number of changes in the curvature sign and on the amplitude (maximum distance of the curve from the underlying chord) of the nerve curves. These measures were evaluated according to their capability to reproduce the expert classification of images into three groups of tortuosity (low, mid, and high). This classification was also compared with measures provided by other methods proposed in the literature to evaluate nerve tortuosity. RESULTS. Among all considered methods, the one proposed herein allows a minimum of classification errors (only 2 in 30 images) and the highest Krippendorff concordance coefficient (0.96). Furthermore, it is the only one that can provide a significant difference (P < 0.01) between all pairs of tortuosity classes. CONCLUSIONS. The results provided by the proposed system confirmed its ability to perform a clinically significant evaluation of corneal nerve tortuosity.

Tortuosity classification of corneal nerves images using a multiple-scale-multiple-window approach

“…We observe that similar studies [5,12] report tests with similar or smaller number of images. The images were chosen so to balance image numbers in each of the 4 tortuosity classes considered.…”

“…[5] were the first to propose an objective, semi-automated method for quantifying sub-basal nerve tortuosity. [12] adapted the algorithm in [10] to work with corneal nerve images; the algorithm performed well with a data set of 30 images divided into 3 classes by a single ophthalmologist. We present a novel approach to the estimation of corneal fibre tortuosity, suitable for retinal vessels as well.…”

Recent Perspectives on Corneal Nerves: A Short Review

“…19,20 It has broadened the scope for non-surgical intervention and cellular examination of live corneas in vivo. [21][22][23][24][25][26] However, despite breakthroughs in imaging techniques, the distribution of corneal nerves is not completely deciphered as yet and the reasons for this lack of understanding is because of the difficulty in obtaining detailed innervations in the different corneal layers since conventional histology requires fresh corneas. Secondly, transmission electron microscopy (TEM) images are restricted to very tiny areas of the corneal surface (0.1 mm 2 ).…”

A fully automatic nerve segmentation and morphometric parameter quantification system for early diagnosis of diabetic neuropathy in corneal images

“…It correctly recognized 80.4% and 83.8% of nerve length, compared with the manually traced nerve length, in control subjects and patients, respectively. Scarpa and et al [36] presented an automatic algorithm to calculate and classify the tortuosity of corneal nerves using a dataset containing 30 corneal sub-basal nerve images. The proposed algorithm depends on the tracing and recognition system for corneal nerves in [35].…”