Artificial intelligence utilising corneal confocal microscopy for the diagnosis of peripheral neuropathy in diabetes mellitus and prediabetes

Preston, Frank G ORCID: 0000-0002-3953-331X, Meng, Yanda ORCID: 0000-0001-7344-2174, Burgess, Jamie ORCID: 0000-0002-7165-6918, Ferdousi, Maryam, Azmi, Shazli, Petropoulos, Ioannis N, Kaye, Stephen ORCID: 0000-0003-0390-0592, Malik, Rayaz A, Zheng, Yalin ORCID: 0000-0002-7873-0922 and Alam, Uazman ORCID: 0000-0002-3190-1122 (2021) Artificial intelligence utilising corneal confocal microscopy for the diagnosis of peripheral neuropathy in diabetes mellitus and prediabetes. DIABETOLOGIA, 65 (3). pp. 457-466.

Access the full-text of this item by clicking on the Open Access link.

Official URL: http://dx.doi.org/10.1007/s00125-021-05617-x

Abstract

<h4>Aims/hypothesis</h4>We aimed to develop an artificial intelligence (AI)-based deep learning algorithm (DLA) applying attribution methods without image segmentation to corneal confocal microscopy images and to accurately classify peripheral neuropathy (or lack of).<h4>Methods</h4>The AI-based DLA utilised convolutional neural networks with data augmentation to increase the algorithm's generalisability. The algorithm was trained using a high-end graphics processor for 300 epochs on 329 corneal nerve images and tested on 40 images (1 image/participant). Participants consisted of healthy volunteer (HV) participants (n = 90) and participants with type 1 diabetes (n = 88), type 2 diabetes (n = 141) and prediabetes (n = 50) (defined as impaired fasting glucose, impaired glucose tolerance or a combination of both), and were classified into HV, those without neuropathy (PN-) (n = 149) and those with neuropathy (PN+) (n = 130). For the AI-based DLA, a modified residual neural network called ResNet-50 was developed and used to extract features from images and perform classification. The algorithm was tested on 40 participants (15 HV, 13 PN-, 12 PN+). Attribution methods gradient-weighted class activation mapping (Grad-CAM), Guided Grad-CAM and occlusion sensitivity displayed the areas within the image that had the greatest impact on the decision of the algorithm.<h4>Results</h4>The results were as follows: HV: recall of 1.0 (95% CI 1.0, 1.0), precision of 0.83 (95% CI 0.65, 1.0), F<sub>1</sub>-score of 0.91 (95% CI 0.79, 1.0); PN-: recall of 0.85 (95% CI 0.62, 1.0), precision of 0.92 (95% CI 0.73, 1.0), F<sub>1</sub>-score of 0.88 (95% CI 0.71, 1.0); PN+: recall of 0.83 (95% CI 0.58, 1.0), precision of 1.0 (95% CI 1.0, 1.0), F<sub>1</sub>-score of 0.91 (95% CI 0.74, 1.0). The features displayed by the attribution methods demonstrated more corneal nerves in HV, a reduction in corneal nerves for PN- and an absence of corneal nerves for PN+ images.<h4>Conclusions/interpretation</h4>We demonstrate promising results in the rapid classification of peripheral neuropathy using a single corneal image. A large-scale multicentre validation study is required to assess the utility of AI-based DLA in screening and diagnostic programmes for diabetic neuropathy.

Item Type:	Article
Uncontrolled Keywords:	Artificial intelligence, Convolutional neural network, Corneal confocal microscopy, Deep learning algorithm, Diabetic neuropathy, Image segmentation, Ophthalmic imaging, Small nerve fibres
Divisions:	Faculty of Health and Life Sciences Faculty of Health and Life Sciences > Institute of Life Courses and Medical Sciences
Depositing User:	Symplectic Admin
Date Deposited:	06 Dec 2021 08:20
Last Modified:	18 Jan 2023 21:23
DOI:	10.1007/s00125-021-05617-x
Open Access URL:	https://link.springer.com/article/10.1007%2Fs00125...
Related URLs:	Author Publisher
URI:	https://livrepository.liverpool.ac.uk/id/eprint/3144516