Cytoarchitecture of epithelial inflammatory infiltration indicates the aetiology of infectious keratitis

Smędowski, Adrian; Tarnawska, Dorota; Orski, M; Wróblewska‐Czajka, Ewa; Aragona, Pasquale; Wylęgała, Edward

doi:10.1111/aos.13363

Cited by 15 publications

(21 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The accuracy of transfer model was comparable to that of experienced ophthalmologist and better than under-experienced ophthalmologist. It has been reported that activated DCs and inflammatory cells are important biomarkers for monitoring inflammatory activity and clinical severity of corneal diseases [ 7 , 8 ]. Our DTL models proposed automated solutions for the evaluation of corneal inflammatory cellular components, which may support under-experienced ophthalmologists in decision making regarding the management of corneal diseases.…”

Section: Discussionmentioning

confidence: 99%

“…In vivo confocal microscopy (IVCM) enables noninvasive analysis of different corneal layers in exquisite detail and allows in vivo detection of even subtle microstructural changes in pathological states [ 3 , 4 ]. IVCM image analysis reveals dendritic cells (DCs) activation and inflammatory cells infiltration in pathologic and infectious conditions such as dry eye [ 5 , 6 ], infectious keratitis of various aetiologies [ 7 , 8 ], and contact lens-induced corneal changes [ 9 ]. The inflammatory cellular components are considered as excellent indicators of inflammatory activity and clinical severity [ 5 , 8 ].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A deep transfer learning framework for the automated assessment of corneal inflammation on in vivo confocal microscopy images

Xu¹,

Qin²,

He³

et al. 2021

PLoS ONE

View full text Add to dashboard Cite

Purpose Infiltration of activated dendritic cells and inflammatory cells in cornea represents an important marker for defining corneal inflammation. Deep transfer learning has presented a promising potential and is gaining more importance in computer assisted diagnosis. This study aimed to develop deep transfer learning models for automatic detection of activated dendritic cells and inflammatory cells using in vivo confocal microscopy images. Methods A total of 3453 images was used to train the models. External validation was performed on an independent test set of 558 images. A ground-truth label was assigned to each image by a panel of cornea specialists. We constructed a deep transfer learning network that consisted of a pre-trained network and an adaptation layer. In this work, five pre-trained networks were considered, namely VGG-16, ResNet-101, Inception V3, Xception, and Inception-ResNet V2. The performance of each transfer network was evaluated by calculating the area under the curve (AUC) of receiver operating characteristic, accuracy, sensitivity, specificity, and G mean. Results The best performance was achieved by Inception-ResNet V2 transfer model. In the validation set, the best transfer system achieved an AUC of 0.9646 (P<0.001) in identifying activated dendritic cells (accuracy, 0.9319; sensitivity, 0.8171; specificity, 0.9517; and G mean, 0.8872), and 0.9901 (P<0.001) in identifying inflammatory cells (accuracy, 0.9767; sensitivity, 0.9174; specificity, 0.9931; and G mean, 0.9545). Conclusions The deep transfer learning models provide a completely automated analysis of corneal inflammatory cellular components with high accuracy. The implementation of such models would greatly benefit the management of corneal diseases and reduce workloads for ophthalmologists.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

A deep transfer learning framework for the automated assessment of corneal inflammation on in vivo confocal microscopy images

Xu¹,

Qin²,

He³

et al. 2021

PLoS ONE

View full text Add to dashboard Cite

show abstract

“…1B–F ). Inflammatory round cells were described as small (<10 µm 13 ), hyperreflective, oval-shaped cells with no visible fibrils, ring shape, or double-walled structure ( Fig. 1G ).…”

Section: Methodsmentioning

confidence: 99%

Corneal Changes in Acanthamoeba Keratitis at Various Levels of Severity: An In Vivo Confocal Microscopic Study

Wei

Cao

Wang

et al. 2021

Trans. Vis. Sci. Tech.

View full text Add to dashboard Cite

Purpose To investigate the relationship between Acanthamoeba cysts and inflammatory cells in Acanthamoeba keratitis (AK) by in vivo confocal microscopy (IVCM). Methods A case-control study included 30 patients with AK and 20 normal controls. The severity of the AK was divided into mild, moderate, and severe. The central cornea and four standard quadrants of the peripheral cornea were imaged by IVCM. The cyst infiltration and dendritic cell (DC) density and maturity (size, length, field, and number of dendrites) were quantified. The relationship between clinical severity, cyst density, and DC alterations was characterized by Spearman correlation analysis. Results The maximum cyst density in the mild, moderate, and severe groups was 31.3 cysts/mm 2 (17.2–32.8), 62.5 cysts/mm 2 (59.3–103.1), and 162.5 cysts/mm 2 (65.6–215.6), respectively. Compared to normal participants, a significant increase in the mean corneal DC density was detected in patients with AK (290.2 ± 97.0 vs. 25.3 ± 8.3 cells/mm 2 ; P < 0.001). Patients with AK presented an increase in median DC size (178.3 vs. 63.6 µm 2 /cell, P < 0.001), median DC field (518.1 vs. 256.6 µm 2 /cell, P = 0.008), and median DC dendrite length (42.3 vs. 14.7 µm/cell, P < 0.001). Increased AK severity was correlated with an increase in cyst density, DC size, and dendrite length (all P < 0.05). An increase in cyst density was significantly correlated with an increase in DC density (β = 0.484, P = 0.026) and DC size (β = 0.557, P = 0.009). Conclusions Cyst density and depth of infiltration as well as maturity of the surrounding DC increased significantly with the severity of AK. Translational Relevance Quantitative analysis of cyst density and DC maturity may provide a new method of evaluating the severity of AK.

show abstract

“…Confocal microscopy is a novel non-invasive corneal imaging examination method [9][10][11]. Due to high resolution and magnification, it reveals the corneal changes at the cellular level under three-dimensional space and real-time conditions; similar results could be obtained in the cloudy corneal tissue, making it a commonly used tool for the clinical study of corneal lesions [9][10][11][12][13][14].…”

Section: Introductionmentioning

confidence: 99%

“…Confocal microscopy plays a major role in evaluating keratitis and corneal nerves [9][10][11][12][13][14], but there is only one study about its use for AACC which focused on Langerhans cells and keratocytes [15]. Our study aimed to investigate corneal epithelial cells, subepithelial nerve fiber plexus, stromal cells, and endothelial cells of the AACC and fellow eyes and healthy controls using confocal microscopy.…”

Section: Introductionmentioning

confidence: 99%

Corneal confocal microscopic characteristics of acute angle-closure crisis

Wang

Yang

et al. 2022

BMC Ophthalmol

View full text Add to dashboard Cite

Background To investigate characteristics of the acute angle-closure crisis (AACC) and fellow eyes using confocal microscopy. Methods Unilateral AACC patients hospitalized at the Xi’an People’s Hospital from October 2017 to October 2020 were recruited in this cross-sectional study. Age-matched participants scheduled for cataract surgery were enrolled as a healthy control group. Corneal epithelial cells, subepithelial nerve fiber plexus, stromal cells, and endothelial cells were examined by confocal and specular microscopy. Results This study enrolled 41 unilateral AACC patients (82 eyes) and 20 healthy controls (40 eyes). Confocal microscopy revealed that the corneal nerve fiber density, corneal nerve branch density and corneal nerve fiber length were reduced significantly in AACC eyes. The stromal cells were swollen and the size of the endothelial cells was uneven with the deposition of punctate high-reflective keratic precipitate on the surface. In severe cases, the cell volume was enlarged, deformed, and fused. The corneal subepithelial nerve fiber, stromal layer, and endothelial layer were unremarkable in the fellow eyes, and the density of the endothelial cells was 2601 ± 529 cells/mm2, which was higher than 1654 ± 999 cells/mm2 in AACC eyes (P < 0.001). Corneal edema prevented the examination of 17 eyes using specular microscopy and in only four eyes using confocal microscopy. There were no significant differences in endothelial cell density between confocal and specular microscopy in the AACC eyes (P = 0.674) and fellow eyes (P = 0.247). The hexagonal cell ratio reduced significantly (P < 0.001), and average cell size and coefficient of variation of the endothelial cells increased significantly compared with fellow eyes (P < 0.001, P = 0.008). Conclusions AACC eye showed decreased density and length of corneal subepithelial nerve fiber plexus, activation of stromal cells, increased endothelial cell polymorphism, and decreased density.

show abstract

Cytoarchitecture of epithelial inflammatory infiltration indicates the aetiology of infectious keratitis

Cited by 15 publications

References 39 publications

A deep transfer learning framework for the automated assessment of corneal inflammation on in vivo confocal microscopy images

A deep transfer learning framework for the automated assessment of corneal inflammation on in vivo confocal microscopy images

Corneal Changes in Acanthamoeba Keratitis at Various Levels of Severity: An In Vivo Confocal Microscopic Study

Corneal confocal microscopic characteristics of acute angle-closure crisis

Contact Info

Product

Resources

About