Corneal nerves and dendritic cells are increasingly being visualised to serve as clinical parameters in the diagnosis of ocular surface diseases using intravital confocal microscopy. In this review, different methods of image analysis are presented. The use of deep learning algorithms, which enable automated pattern recognition, is explained in detail using our own developments and compared with other established methods.
Hornhautnerven und dendritische Zellen werden zunehmend bei der Diagnostik von Erkrankungen der Augenoberfläche als klinische Parameter mittels intravitaler Konfokalmikroskopie dargestellt. In dieser Übersichtsarbeit werden unterschiedliche Verfahren der Bildauswertung dargestellt. Die Verwendung von Deep-Learning-Algorithmen, die eine automatisierte Mustererkennung ermöglichen, wird anhand eigener Entwicklungen detailliert erläutert und mit anderen etablierten Verfahren verglichen.

OBJECTIVE: To compare, between Alzheimer\'s disease (AD) patients and healthy individuals, corneal subbasal nerve plexus (CSNP) parameters and corneal sensitivities.
METHODS: Twenty-two patients who were followed up with Alzheimer\'s disease (Alzheimer\'s group) and 18 age- and gender-matched healthy individuals (control group) were included in this cross-sectional study. CSNP parameters, including nerve fiber length (NFL), nerve fiber density (NFD), and nerve branch density (NBD), were evaluated using in vivo confocal microscopy. Corneal sensitivity was evaluated using a Cochet-Bonnet esthesiometer. The results were compared between the two groups.
RESULTS: In the Alzheimer\'s group, NFL was 12.2 (2.4) mm/mm2, NFD was 12.5 [3.1] fibers/mm2, and NBD was 29.7 [9.37] branches/mm2. In the control group, NFL was 16.5 (2.0) mm/mm2, NFD was 25.0 [3.13] fibers/mm2, and NBD was 37.5 [10.9] branches/mm2. All three parameters were significantly lower in the Alzheimer\'s group compared to the control group (p < 0.001, p < 0.001, and p = 0.001, respectively). Similarly, corneal sensitivity was significantly lower in the Alzheimer\'s group (55.0 [5.0] mm) compared to the control group (60.0 [5.0] mm) (p < 0.001).
CONCLUSIONS: We determined that, in AD, corneal sensitivity decreases significantly, in parallel with the decrease in corneal nerves. Changes in the corneal nerve plexus and a decrease in corneal sensitivity may be used in the early diagnosis and follow-up of AD. In addition, ocular surface problems secondary to these changes should also be kept in mind.

There has been a growing application of in vivo confocal microscopy (IVCM) in the examination of corneal microstructure, including different corneal layers and corneal nerve fibers in health and in pathological conditions. Corneal nerves forming the sub-basal nerve plexus (SBNP) beneath the corneal basal epithelial cell layer in particular have been intensively researched in health and disease as a marker for corneal neurophysioanatomical and degenerative changes. One intriguing feature in the SBNP that is found inferior to the corneal apex, is a whorl-like pattern (or vortex) of nerves, which represents an anatomical landmark. Evidence has indicated that the architecture of this \'whorl region\' is dynamic, changing with time in healthy individuals but also in disease conditions such as in diabetic neuropathy and keratoconus. This review summarizes the known information regarding the characteristics and significance of the whorl region of nerves in the corneal SBNP, as a potential area of high relevance for future disease monitoring and diagnostics.

During life up to 70% of aniridia subjects develop aniridia-associated keratopathy (AAK). AAK is characterized by limbal stem cell insufficiency, impaired corneal epithelial cell differentiation and abnormal cell adhesion, which leads to centripetal spreading vascularization, conjunctivalization, and thickening of the cornea. Our aim was to examine the subbasal nerve plexus and central corneal stromal microstructure in subjects with congenital aniridia, using in vivo confocal laser scanning microscopy CLSM.
31 eyes of 18 patients (55.6% males, mean age: 25.22 ± 16.35 years) with congenital aniridia and 46 eyes of 29 healthy subjects (41.4% males, mean age 30 ± 14.82 years) were examined using the Rostock Cornea Module of Heidelberg Retina Tomograph-III. At the subbasal nerve plexus, corneal nerve fiber density (CNFD), corneal nerve fiber length (CNFL), corneal total branch density (CTBD), and corneal nerve fiber width (CNFW) were analyzed using ACCMetrics software. Keratocyte density in the anterior, middle and posterior stroma was assessed manually.
The CNFD (2.02 ± 4.08 vs 13.99 ± 6.34/mm2), CNFL (5.78 ± 2.68 vs 10.56 ± 2.82 mm/mm2) and CTBD (15.08 ± 15.62 vs 27.44 ± 15.05/mm2) were significantly lower in congenital aniridia subjects than in controls (p < 0.001 for all). CNFW was significantly higher in aniridia subjects than in controls (0.03 ± 0.004 vs 0.02 ± 0.003 mm/mm2) (p = 0.003). Keratocyte density was significantly lower in all stromal layers of aniridia subjects than in controls (p < 0.001 for all). Stromal alterations included confluent keratocytes, keratocytes with long extensions and hyperreflective dots between keratocytes in aniridia.
Decrease in CNFD, CNFL, and CTBD, as well as increase in CNFW well refer to the congenital aniridia-associated neuropathy. The decreased keratocyte density and the stromal alterations may be related to an increased cell death in congenital aniridia, nevertheless, stromal changes in different stages of AAK have to be further analyzed in detail.

OBJECTIVE: Migraine is a chronic neurovascular disease that affects the trigeminovascular system. The purpose of this study was to evaluate corneal subbasal nerve fibers, dendritic cells and to measure tear film parameters in migraine.
METHODS: 87 eyes of 44 patients suffering from migraine with a mean age of 33.23 ± 11.41 years were included in our study. 25 age-matched controls (mean age of 30.16 ± 12.59 years; P = 0.162) were recruited. The corneal subbasal plexus and the dendritic cells (DC) were analyzed using in vivo confocal microscopy (Heidelberg Retina Tomograph II Rostock Cornea Module; Heidelberg Engineering GmbH), and the tear film was imaged using LacryDiag (Quantel Medical, France).
RESULTS: Regarding the subbasal nerve fibers of the cornea, none of the examined parameters differed significantly in migraine patients from controls. We found a significant increase in the corneal DC density (P < 0.0001) and DC area (P < 0.0001) in migraine patients compared to healthy volunteers. DC density showed a positive correlation with the monthly attack frequency (r = 0.32, P = 0.041) and the DC area a negative correlation with corneal nerve branch density (r = -0.233, P = 0.039), nerve fiber length (r = -0.232, P = 0.04) and total branch density (r = -0.233, P = 0.039). Using LacryDiag a significant loss of Meibomian gland area could be detected on the superior eyelid (P = 0.005) in migraine.
CONCLUSIONS: Our results suggest the presence of neuroinflammation in the cornea of migraine patients affecting the peripheral trigeminal system. Dendritic cells surrounding the subbasal plexus may be involved in the activation and modulation of pain in migraine.

暂无摘要。

Surgeons dissect carefully in the medial third of the supraorbital rim to preserve the supraorbital nerve (SON) during surgical forehead rejuvenation. However, the anatomic variations of SON exit from the frontal bone have been researched in cadaver or imaging studies. In this study, we report a variation in the lateral branch of SON observed in an endoscopic view during forehead lifts. A retrospective review of 462 patients who underwent endoscopy-assisted forehead lifts between January 2013 and April 2020 was performed. Data, including the location, number, and form of the exit point and thickness of SON and its lateral branch variant, were recorded and reviewed intraoperatively, utilizing high-definition endoscopic assistance. Thirty-nine patients and 51 sides were included, and all patients were female, with a mean age of 44.53 (18-75) years. This nerve exited a foramen in the frontal bone ~8.82 ± 2.79 mm lateral to SON and ~1.89 ± 1.34 mm from the supraorbital margin vertically. Observed thickness variations of the lateral branch of SON included 20 small, 25 medium, and 6 large nerves. This study revealed various positional and morphologic variations of the lateral branch of SON in an endoscopic view. Thus, surgeons can be alerted of the anatomic variations of SON and establish careful dissection during procedures. In addition, the findings of this study will be useful in planning nerve blocks, filler injections, and migraine treatments in the supraorbital region.

Sensory alterations of the upper eyelid skin and eyelashes are frequently encountered after upper eyelid surgery. The objective of this study was to provide information on the exact course and distribution of sensory nerve fibers through the anatomic planes in the upper eyelid.
Ten formalin-fixed hemifaces were dissected. The nerve branches of the ophthalmic nerve in the upper eyelid were traced in an anterograde fashion.
A total of 151 nerve fibers were recorded during dissection. The infratrochlear, supratrochlear, supraorbital, and lacrimal nerve contributed each to both the upper eyelid skin innervation and the upper eyelid rim plexus in different distribution patterns. The mean distance from the eyelid margin at which nerve fibers pierced from preseptal into the orbicularis muscle was 14 ± 1.1 mm for nerve fibers targeting the eyelid dermis and 3.7 ± 1.2 mm for nerve fibers targeting the eyelid rim plexus ( p < 0.001). The mean intraorbicular course of nerve fibers was 3 mm (0-17; standard deviation 4.1). The mean distance from the eyelid margin at which nerve fibers pierced from the orbicularis muscle into the preorbicular plane was 10 ± 1 mm for nerve fibers innervating the eyelid dermis and 1.3 ± 0.8 mm for nerve fibers innervating the eyelid rim plexus ( p < 0.001). The mean distance of the preorbicular course of nerve fibers was 2 mm (0-15; standard deviation 3.6).
Based on the findings, a certain degree of postoperative eyelid skin numbness is inevitable while eyelash innervation may be spared in upper blepharoplasty.

OBJECTIVE: Small-fiber neuropathy (SFN) is known to be associated with Sjögren disease (SjD), and in vivo corneal confocal microscopy can identify features compatible with SFN. Here, we performed a descriptive study to identify features of SFN of the corneal subbasal nerve plexus using in vivo confocal microscopy.
METHODS: We recruited 10 participants from the Sjögren\'s International Collaborative Clinical Alliance (SICCA), 1 new participant (in an effort to expand the SICCA cohort), and 22 healthy controls. All participants underwent slit-lamp examination and in vivo confocal microscopy of the central corneal subbasal nerve plexus centered about the central whorl to create a 30-image montage. Each image was analyzed with automated software (ACCmetrics, Manchester, United Kingdom) to produce 7 nerve metrics. We performed t-tests and age-adjusted regressions to make comparisons of nerve metrics between participants with SjD and healthy controls.
RESULTS: Most nerve metrics were significantly lower in participants with SjD compared with healthy controls. The mean corneal nerve fiber density was found to be 3.5 mm/mm 2 in participants with SjD compared with 10.6 mm/mm 2 in healthy controls (95% confidence interval, -8.4 to -0.93; P = 0.02). Within the 11 participants with SjD, 22 eyes were analyzed on confocal microscopy, and 16 of those eyes (from 9 individuals) did not have an identifiable central whorl. Within the 22 healthy controls, 22 eyes (right eye alone) were analyzed on confocal microscopy, and 21 of those eyes had an identifiable central whorl.
CONCLUSIONS: SjD exhibits lower corneal nerve metrics compared with healthy controls. These findings suggest that features compatible with SFN can distinguish SjD from healthy controls and may serve as a potential novel biomarker in identifying SjD.

Sensory nerve tension and gliding-layer mobility in the brow may be significant factors affecting postoperative brow level in an endoscopic brow lift, yet they have rarely been studied.
To investigate the effects of sensory nerve tension and gliding-layer mobility, the following measurements were performed alongside the endoscopic brow lift in 50 fresh cadaveric hemifaces: amount of brow elevation, critical lifting amount (as sensory nerves became tense), laxity of sensory nerve courses, and mobility of brow-gliding layers. The sensory nerve situations in the subperiosteal and subgaleal dissections were also observed.
Supraorbital nerve tension limited the cephalic advancement of the forehead flap. The mean elevation of the brow was 5.8 ± 1 mm (range, 3.5 to 8.6 mm). The mean critical lifting amount was 5.3 ± 1.1 mm (range, 4.0 to 7.3 mm). The mean amount of laxity in the supraorbital nerve (the permissible amount of lift) was 4.1 ± 0.9 mm (range, 2.5 to 5.5 mm). The galeal fat pad was responsible for 60% of brow mobility. The sensory nerve was more protected by a subgaleal dissection in the brow and inferior forehead and by a subperiosteal dissection in the middle and upper forehead.
Cephalic movement of the forehead flap is limited by supraorbital nerve tension. The permitted lifting amount varies from 2.5 to 5.5 mm. Gliding-layer mobility in the brow offsets the postoperative amount of cephalic advancement of the forehead flap. Consideration of supraorbital nerve tension and gliding-layer mobility is recommended to obtain an optimal brow level in endoscopic brow lifts.