Development and validation of a deep learning model for diagnosing neuropathic corneal pain via in vivo confocal microscopy

Abstract Neuropathic corneal pain (NCP) is an underdiagnosed ocular disorder caused by aberrant nociception and hypersensitivity of corneal nerves, often resulting in chronic pain and discomfort even in the absence of noxious stimuli. Recently, microneuromas (aberrant growth and swelling of the corn...

Full description

Saved in:
Bibliographic Details
Main Authors: Neslihan Dilruba Koseoglu, Eric Chen, Rudraksh Tuwani, Benjamin Kompa, Stephanie M. Cox, M. Cuneyt Ozmen, Mina Massaro-Giordano, Andrew L. Beam, Pedram Hamrah
Format: Article
Language:English
Published: Nature Portfolio 2025-05-01
Series:npj Digital Medicine
Online Access:https://doi.org/10.1038/s41746-025-01577-3
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850273344067207168
author Neslihan Dilruba Koseoglu
Eric Chen
Rudraksh Tuwani
Benjamin Kompa
Stephanie M. Cox
M. Cuneyt Ozmen
Mina Massaro-Giordano
Andrew L. Beam
Pedram Hamrah
author_facet Neslihan Dilruba Koseoglu
Eric Chen
Rudraksh Tuwani
Benjamin Kompa
Stephanie M. Cox
M. Cuneyt Ozmen
Mina Massaro-Giordano
Andrew L. Beam
Pedram Hamrah
author_sort Neslihan Dilruba Koseoglu
collection DOAJ
description Abstract Neuropathic corneal pain (NCP) is an underdiagnosed ocular disorder caused by aberrant nociception and hypersensitivity of corneal nerves, often resulting in chronic pain and discomfort even in the absence of noxious stimuli. Recently, microneuromas (aberrant growth and swelling of the corneal nerve endings) detected using in vivo confocal microscopy (IVCM) have emerged as a promising biomarker for NCP. However, this process is time-intensive and error-prone, limiting its clinical use and availability. In this work, we present a new NCP screening system based on a deep learning model trained to detect microneuromas using a multisite dataset with a combined total of 103,168 IVCM images. Our model showed excellent discriminative ability detecting microneuromas (AuROC: 0.97) and the ability to generalize to data from a new institution (AuROC: 0.90). Additionally, our pipeline provides an uncertainty quantification mechanism that allows it to communicate when its predictions are reliable, further increasing its clinical relevance.
format Article
id doaj-art-ff62926e0aba492e911f742483f8c7d0
institution OA Journals
issn 2398-6352
language English
publishDate 2025-05-01
publisher Nature Portfolio
record_format Article
series npj Digital Medicine
spelling doaj-art-ff62926e0aba492e911f742483f8c7d02025-08-20T01:51:31ZengNature Portfolionpj Digital Medicine2398-63522025-05-01811810.1038/s41746-025-01577-3Development and validation of a deep learning model for diagnosing neuropathic corneal pain via in vivo confocal microscopyNeslihan Dilruba Koseoglu0Eric Chen1Rudraksh Tuwani2Benjamin Kompa3Stephanie M. Cox4M. Cuneyt Ozmen5Mina Massaro-Giordano6Andrew L. Beam7Pedram Hamrah8Center for Translational Ocular Immunology and Cornea Service, New England Eye Center, Department of Ophthalmology, Tufts Medical Center, Tufts University School of MedicineHarvard-MIT Program in Health Sciences and TechnologyDepartment of Epidemiology, Harvard T.H. Chan School of Public HealthDepartment of Biomedical Informatics, Harvard Medical SchoolCenter for Translational Ocular Immunology and Cornea Service, New England Eye Center, Department of Ophthalmology, Tufts Medical Center, Tufts University School of MedicineCenter for Translational Ocular Immunology and Cornea Service, New England Eye Center, Department of Ophthalmology, Tufts Medical Center, Tufts University School of MedicineScheie Eye Institute, Department of Ophthalmology, University of PennsylvaniaDepartment of Epidemiology, Harvard T.H. Chan School of Public HealthCenter for Translational Ocular Immunology and Cornea Service, New England Eye Center, Department of Ophthalmology, Tufts Medical Center, Tufts University School of MedicineAbstract Neuropathic corneal pain (NCP) is an underdiagnosed ocular disorder caused by aberrant nociception and hypersensitivity of corneal nerves, often resulting in chronic pain and discomfort even in the absence of noxious stimuli. Recently, microneuromas (aberrant growth and swelling of the corneal nerve endings) detected using in vivo confocal microscopy (IVCM) have emerged as a promising biomarker for NCP. However, this process is time-intensive and error-prone, limiting its clinical use and availability. In this work, we present a new NCP screening system based on a deep learning model trained to detect microneuromas using a multisite dataset with a combined total of 103,168 IVCM images. Our model showed excellent discriminative ability detecting microneuromas (AuROC: 0.97) and the ability to generalize to data from a new institution (AuROC: 0.90). Additionally, our pipeline provides an uncertainty quantification mechanism that allows it to communicate when its predictions are reliable, further increasing its clinical relevance.https://doi.org/10.1038/s41746-025-01577-3
spellingShingle Neslihan Dilruba Koseoglu
Eric Chen
Rudraksh Tuwani
Benjamin Kompa
Stephanie M. Cox
M. Cuneyt Ozmen
Mina Massaro-Giordano
Andrew L. Beam
Pedram Hamrah
Development and validation of a deep learning model for diagnosing neuropathic corneal pain via in vivo confocal microscopy
npj Digital Medicine
title Development and validation of a deep learning model for diagnosing neuropathic corneal pain via in vivo confocal microscopy
title_full Development and validation of a deep learning model for diagnosing neuropathic corneal pain via in vivo confocal microscopy
title_fullStr Development and validation of a deep learning model for diagnosing neuropathic corneal pain via in vivo confocal microscopy
title_full_unstemmed Development and validation of a deep learning model for diagnosing neuropathic corneal pain via in vivo confocal microscopy
title_short Development and validation of a deep learning model for diagnosing neuropathic corneal pain via in vivo confocal microscopy
title_sort development and validation of a deep learning model for diagnosing neuropathic corneal pain via in vivo confocal microscopy
url https://doi.org/10.1038/s41746-025-01577-3
work_keys_str_mv AT neslihandilrubakoseoglu developmentandvalidationofadeeplearningmodelfordiagnosingneuropathiccornealpainviainvivoconfocalmicroscopy
AT ericchen developmentandvalidationofadeeplearningmodelfordiagnosingneuropathiccornealpainviainvivoconfocalmicroscopy
AT rudrakshtuwani developmentandvalidationofadeeplearningmodelfordiagnosingneuropathiccornealpainviainvivoconfocalmicroscopy
AT benjaminkompa developmentandvalidationofadeeplearningmodelfordiagnosingneuropathiccornealpainviainvivoconfocalmicroscopy
AT stephaniemcox developmentandvalidationofadeeplearningmodelfordiagnosingneuropathiccornealpainviainvivoconfocalmicroscopy
AT mcuneytozmen developmentandvalidationofadeeplearningmodelfordiagnosingneuropathiccornealpainviainvivoconfocalmicroscopy
AT minamassarogiordano developmentandvalidationofadeeplearningmodelfordiagnosingneuropathiccornealpainviainvivoconfocalmicroscopy
AT andrewlbeam developmentandvalidationofadeeplearningmodelfordiagnosingneuropathiccornealpainviainvivoconfocalmicroscopy
AT pedramhamrah developmentandvalidationofadeeplearningmodelfordiagnosingneuropathiccornealpainviainvivoconfocalmicroscopy