transferGWAS: GWAS of images using deep transfer learning.

MOTIVATION: Medical images can provide rich information about diseases and their biology. However, investigating their association with genetic variation requires non-standard methods. We propose transferGWAS, a novel approach to perform genome-wide association studies directly on full medical images. First, we learn semantically meaningful representations of the images based on a transfer learning task, during which a deep neural network is trained on independent but similar data. Then, we perform genetic association tests with these representations. RESULTS: We validate the type I error rates and power of transferGWAS in simulation studies of synthetic images. Then we apply transferGWAS in a genome-wide association study of retinal fundus images from the UK Biobank. This first-of-a-kind GWAS of full imaging data yielded 60 genomic regions associated with retinal fundus images, of which 7 are novel candidate loci for eye-related traits and diseases. AVAILABILITY AND IMPLEMENTATION: Our method is implemented in Python and available at https://github.com/mkirchler/transferGWAS/. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Rapid Campimetry-A Novel Screening Method for Glaucoma Diagnosis.

One of the most important functions of the retina-the enabling of perception of fast movements-is largely suppressed in standard automated perimetry (SAP) and kinetic perimetry (Goldmann) due to slow motion and low contrast between test points and environment. Rapid campimetry integrates fast motion (=10°/4.7 s at 40 cm patient-monitor distance) and high contrast into the visual field (VF) examination in order to facilitate the detection of absolute scotomas. A bright test point moves on a dark background through the central 10° VF. Depending on the distance to the fixation point, the test point automatically changes diameter (≈0.16° to ≈0.39°). This method was compared to SAP (10-2 program) for six subjects with glaucoma. Rapid campimetry proved to be comparable and possibly better than 10-2 SAP in identifying macular arcuate scotomas. In four subjects, rapid campimetry detected a narrow arcuate absolute scotoma corresponding to the nerve fiber course, which was not identified as such with SAP. Rapid campimetry promises a fast screening method for the detection of absolute scotomas in the central 10° visual field, with a potential for cloud technologies and telemedical applications. Our proof-of-concept study motivates systematic testing of this novel method in a larger cohort.

Cyclophotocoagulation and cyclocryocoagulation as primary surgical procedures for open-angle glaucoma.

PURPOSE: Comparative study of cyclophotocoagulation (CPC) and cyclocryocoagulation (CCT) as primary surgical procedures in patients with open-angle glaucoma with regard to efficacy and complications. METHODS: In this retrospective cohort study, 184 eyes of 112 patients in whom cyclodestructive surgery was performed as a primary surgical procedure were examined. CPC was performed on 133 eyes and CCT on 51 eyes. A standardised multiple measurement of intraocular pressure (IOP) was performed on all patients preoperatively and at the follow-up examination after an average of 5.5 (1.5-12) months. In addition, the best-corrected visual acuity and the number of antiglaucoma agents were recorded. RESULTS: On average, a reduction in IOP was observed after both of the cyclodestructive procedures (CPC: -1.55 ± 2.50 mmHg, p < 0.05; CCT: -2.33 ± 3.06 mmHg; p < 0.05). The average difference in IOP reduction between the two procedures (0.78 mmHg) proved to be statistically insignificant (p = 0.08). In contrast, greater patient age and higher preoperative IOP values were found to be highly significant influencing factors. In 45 % and 70 % of the patients treated with CPC and CCT, respectively, IOP was reduced by at least 20 %, with no increase in medication or with a reduction in medication of at least one substance with no increase in pressure. CPC and CCT produced an average loss of visual acuity of more than two lines in 10.5 % and 9.8 % of cases, respectively. Permanent hypotension did not occur in any of the cases. CONCLUSIONS: A moderate reduction in IOP is achievable with both procedures, with CCT tending to produce a greater reduction in pressure. The efficacy of primary cyclodestructive procedures increases with increasing patient age and with higher preoperative IOP values. The risk of serious complications can be considered low.

Impact of temporary hyperthermia on corneal endothelial cell survival during organ culture preservation.

To evaluate temporary exposure to hyperthermia for its impact on endothelial cell density of porcine corneas in organ culture medium containing dextran with regards to possible negative influences of high temperatures during the storage and transport of corneal grafts. Four groups of central discs (diameter 8 mm) from the corneas of both eyes in 40 pigs were first organ-cultured (MEM with 6% dextran 500) for 24 h at 32°C. Ten corneas were then exposed to 40°C in group 1, to 42°C in group 2, to 44°C in group 3, and to 50°C in group 4 for 12 h each. The paired corneal discs for all groups were not treated, stored at 32°C and served as controls. After further organ culture of all corneas for 48 h at 32°C to allow regenerative processes, corneal endothelium was stained with Alizarin Red S and examined by light microscopy. The endothelial cell densities were determined on three central images using a system for the automatic estimation of morphometric parameters of corneal endothelium. Exposure for 12 h to 40°C as well as to 42°C induced no endothelial cell loss. Statistical analysis showed no significant difference of the endothelial cell density between corneas exposed to 40°C and 42°C and the control corneas (40°C treatment: 4736 ± 426 cells/mm(2) and control: 4762 ± 344 cells/mm(2), p = 0.74; 42°C treatment: 4240 ± 363 cells/mm(2) and control: 4176 ± 448 cells/mm(2), p = 0.40). Exposure to 44°C and 50°C lead to total necrosis of the endothelial cell layer. Exposure of organ cultured porcine corneas in dextran containing medium up to 42°C for 12 h does not compromise the endothelial cell density in a clinically relevant manner. Temperatures above 42°C, as it might be the case during transports from the cornea bank to the ophthalmic surgeon, must be strictly avoided as they damage the endothelial cell layer.

[Patient education and treatment documentation - law to improve the rights of patients]. / Patientenaufklärung und Behandlungsdokumentation - Gesetz zur Verbesserung der Rechte von Patientinnen und Patienten.

The Law to Improve the Rights of Patients came into force with the promulgation in the German Federal Law Gazette on February 25, 2013. Thus administrations of medical institutions and doctors of all disciplines should acquaint themselves with the statutory regulations and their impact on the daily practice. The present article describes and explains the statutory regulations concerning patient education and treatment documentation.

Corneal intrastromal tissue modeling with the femtosecond laser.

PURPOSE: To determine whether central corneal thinning and flattening can be achieved by intrastromal photodisruption using a femtosecond (fs) laser. SETTING: Institute of Clinical Neuroanatomy, Goethe-University, Frankfurt am Main, Germany. METHODS: Fourteen horizontal, parallel intrastromal cuts were performed on rabbit eyes using a fs laser. Full-grown rabbits (group 1; ten eyes) received bilateral treatment. Growing rabbits (group 2) received unilateral treatment (four eyes). Slit-lamp examination, pachymetry, and keratometry were performed on day 9, 31 and 86 (group 1) or on day 12, 29, 69, 176 and 318 (group 2) after surgery. RESULTS: Nine days after treatment, corneal swelling was present and a slight increase of mean corneal thickness (group 1: +4.40 ± 5.56 µm) as well as a steeper mean corneal curvature (group 1: -0.18 ± 0.02 mm) were observed. In contrast, 1 month after tissue photodisruption corneas showed an average decrease of thickness (group 2: -21.0 ± 2.5 µm). By 6 months post-treatment, a further decrease (group 2: -36.3 ± 6.9 µm) was seen that remained stable for the rest of the observation period. At 176 days post-treatment, a decrease of corneal curvature (group 2: -0.21 ± 0.10 mm) was found. Slit-lamp examination revealed a transparent cornea. At the site of intrastromal photodisruption a narrow band of increased reflectivity could be detected. CONCLUSIONS: Corneal thinning can be reliably achieved using intrastromal tissue modeling with a fs laser. Tissue modeling was accompanied by a transient opacity and irregularity of the corneal surface.

A system for the automatic estimation of morphometric parameters of corneal endothelium in alizarine red-stained images.

BACKGROUND/AIMS A computer program for the automatic estimation of endothelium morphometric parameters (cell density, pleomorphism, polymegethism) in alizarine red-stained images is presented and evaluated. METHODS Images of corneal endothelium from 30 porcine eyes stained with alizarine red were acquired with an optical microscope and saved as grey-level digital images. Each image was first pre-processed for luminosity correction and contrast enhancement. An artificial neural network was used to classify all pixels as cell contour or cell body pixels. The segmented cell contours were then used to obtain estimates of morphometric parameters. The central area was assessed and the mean area per cornea was 0.54+/-0.07 mm(2). The whole system was implemented as a computer program using the Matlab language. Estimated parameters were compared with the corresponding values derived from manual contour detection on the same images used for the automatic estimation. RESULTS For the 30 images in our dataset, the mean differences for automatic versus manual parameters were -12+/-52 (range -103 to +145) cells/mm(2) for density, 0.5+/-2.6% (range -5.6 to +5.6%) for pleomorphism and -0.7+/-1.9% (range -4.1 to +2.8%) for polymegethism. CONCLUSION The evaluation of the automatic system on 30 images from porcine eyes confirmed its ability to estimate reliably morphometric parameters with respect to parameter values derived by manual analysis.

Intrastromal keratotomy with femtosecond laser avoids profibrotic TGF-beta1 induction.

PURPOSE: To examine expression of the profibrotic cytokine TGF-beta1 after selective intrastromal corneal injury with the use of a femtosecond laser. METHODS: Rabbits underwent monocular intrastromal keratotomy at a preoperatively determined corneal depth of 160 to 200 mum with the use of a femtosecond laser. Femtosecond laser-induced TGF-beta1 expression was compared in nonoperated control eyes and eyes treated with photorefractive keratectomy (PRK). Follow-up examinations were performed 1, 3, 7, and 28 days after surgery. TGF-beta1 protein was identified by immunofluorescence labeling. With the use of laser-capture microdissection, epithelial, stromal, and endothelial cell layers were collected, and changes in TGF-beta1 mRNA expression were quantified with quantitative RT-PCR. RESULTS: TGF-beta1 mRNA and protein expression did not significantly increase after intrastromal femtosecond laser keratotomy. In contrast, TGF-beta1 was induced in corneal epithelial and stromal cells after PRK and showed up to 23-fold higher TGF-beta1 mRNA levels compared with control corneas. The increase of TGF-beta1 mRNA levels after PRK was accompanied by increased TGF-beta1 protein production. CONCLUSIONS: Isolated stromal injury with a femtosecond laser does not result in induction of the profibrotic cytokine TGF-beta1. Because TGF-beta1 has been implicated in a fibrotic response of the corneal stroma to injury, absence of TGF-beta1 induction argues for a favorable wound-healing response. These findings support highly selective intrastromal procedures in refractive surgery.

Laser in situ keratomileusis following the implantation of iris-fixated phakic intraocular lenses.

PURPOSE: This study aims to assess the safety, efficacy, predictability, stability and complications associated with laser in situ keratomileusis (LASIK) for the correction of residual refractive errors following implantation of iris-fixated phakic intraocular lenses (pIOLs). METHODS: In this retrospective interventional case series of 92 eyes, an iris-fixated pIOL (Artisan) was implanted to correct myopia and myopic astigmatism. In 10 eyes, a residual refractive error was treated using LASIK. Visual acuity testing, subjective refraction, slitlamp examination and tonometry were all performed preoperatively, then 3 months following pIOL implantation and 1 year after LASIK treatment. The endothelial cell density was measured both prior to pIOL implantation and following LASIK therapy. RESULTS: Comparison of preoperative and postoperative best-corrected visual acuity revealed that none of the combined operated eyes forfeited > or =2 lines of visual acuity after LASIK. Uncorrected visual acuity for all 10 eyes after LASIK was > or =0.8. LASIK made a 61.5% mean reduction in astigmatism possible. Compared to the initial examination, the mean endothelial cell loss after LASIK treatment was 4.1%. CONCLUSIONS: For correction of residual refractive errors following iris-fixated phakic IOL implantation, LASIK appears to be a safe and effective procedure. Larger numbers of patients are required to verify this conclusion.

Influence of temporary hypothermia on corneal endothelial cell density during organ culture preservation.

PURPOSE: To evaluate temporary exposure to hypothermia for its effects on endothelial cell density of porcine corneas in dextran containing organ-culture medium, with regard to possible negative influences of low temperatures during the transport of corneal grafts. METHODS: Two groups of central discs from pig corneas (diameter 8 mm) were first organ-cultured (MEM with 6% dextran 500) for 24 hours at 32 degrees C. Twelve corneas were exposed to 4 degrees C in group 1 for 12 hours and to 21 degrees C in group 2 for 48 hours each. The paired corneal discs were not treated, and served as controls. After further organ culture of all corneas for 48 hours at 32 degrees C to allow regenerative processes, corneal endothelium was stained with alizarin red S and examined by light microscopy. The endothelial cell densities were determined manually on three central images. RESULTS: Exposure for 12 hours to 4 degrees C as well as for 48 hours to 21 degrees C induced an endothelial cell loss of 0.3% and 1.8% respectively. Statistical analysis showed no significant difference (p = 0.680) of the endothelial cell density between corneas exposed to 4 degrees C and the control corneas (4166 +/- 389 cells/mm(2) and 4177 +/- 407 cells/mm(2) respectively). Despite the minor cell loss, the difference of the endothelial cell density between corneas exposed to 21 degrees C and the control corneas (4085 +/- 260 cells/mm(2) and 4159 +/- 312 cells/mm(2) respectively) was statistically significant (p = 0.025). CONCLUSIONS: Exposure of organ-cultured porcine corneas in dextran containing medium to 4 degrees C for 12 hours and 21 degrees C for 48 hours does not compromise the endothelial cell density of donor corneas in a clinically relevant manner. A storage of corneal grafts at temperatures down to 4 degrees C for 12 hours, as might be the case during transport from the cornea bank to the ocular surgeon, does not seem to damage the endothelial cell layer.

Corneal femtosecond laser keratotomy results in isolated stromal injury and favorable wound-healing response.

PURPOSE: To examine the corneal repair response after intrastromal femtosecond (fs) laser keratotomy. METHODS: Twelve rabbits underwent monocular intrastromal keratotomy performed with an fs laser at a preoperatively determined corneal depth of 160 to 200 microm. The fs laser-induced corneal repair response was compared with that of nonoperated control eyes and eyes treated with photorefractive keratectomy (PRK). Follow-up examinations were performed 1, 3, 7, and 28 days after surgery. Corneas were evaluated using slit lamp, in vivo confocal microscopy, and light microscopy. The extracellular matrix components fibronectin and tenascin were located using immunofluorescence staining. Anti-Thy-1 and anti-alpha-SMA antibodies and phalloidin were used to identify repair fibroblasts. Cell proliferation and nuclear DNA fragmentation were detected using an anti-Ki-67 antibody and the TUNEL assay, respectively. RESULTS: Intrastromal fs keratotomy resulted in a hypocellular stromal scar discernible as a narrow band of increased reflectivity on slit lamp examination. Deposition of fibronectin and tenascin as well as death and subsequent proliferation of keratocytes were observed. No differentiation of keratocytes into Thy-1- or alpha-SMA-positive fibroblasts could be detected. In contrast, after PRK, which causes epithelial and stromal wounding, all markers for repair fibroblasts were found in subepithelial stromal layers. On slit lamp examination, a fibrotic scar and a corneal haze were revealed. CONCLUSIONS: Isolated stromal injury using an fs laser avoids epithelial injury and is associated with a favorable wound-healing response preserving corneal transparency. Thus, fs laser keratotomy is a highly selective laser treatment that can be useful for the treatment of refractive errors.

Endothelial cell density in porcine corneas after exposure to hypotonic solutions.

PURPOSE: To evaluate exposure to sucrose solution (1.8%) and hypotonic balanced salt solution (BSS) for its effects on endothelial cell density of porcine corneas. METHODS: Two groups of central discs from pig corneas were organ-cultured for 24 h. Twelve corneas per group were exposed to sucrose solution (1.8%) or hypotonic BSS for 4 min each. The paired corneal discs were not treated and served as controls. After further organ culture with and without dextran for 48 h, corneal endothelium was stained with alizarin red and examined by light microscopy. The endothelial cell densities were determined manually on three central images. RESULTS: The endothelial cell density differed significantly between corneas exposed to sucrose and the control corneas (3982+/-382 cells/mm(2) and 4360+/-331 cells/mm(2) respectively, and 3876+/-364 cells/mm(2) versus 4374+/-168 cells/mm(2) respectively with 6% dextran). In contrast, the endothelial cell density did not differ significantly between corneas exposed to hypotonic BSS and the control corneas (4374+/-296 cells/mm(2) and 4317+/-193 cells/mm(2) respectively, and 4348+/-151 cells/mm(2) versus 4426+/-175 cells/mm(2), respectively with 6% dextran). CONCLUSIONS: Exposure to 1.8% sucrose for 4 min induces a significant endothelial cell loss of 10% on average, whereas exposure to hypotonic BSS did not significantly influence the endothelial cell density.

Correlation between clinical in vivo confocal microscopic and ex vivo histopathologic findings of Salzmann nodular degeneration.

PURPOSE: To report clinical, in vivo confocal microscopy and ex vivo histopathologic findings of Salzmann nodular degeneration (SND). METHODS: A 48-year-old woman with symptoms of ocular irritation and decreased visual acuity caused by SND in both eyes was treated by corneal scraping and phototherapeutic keratectomy (PTK). Slit-lamp biomicroscopy, in vivo confocal microscopy, ex vivo light microscopy, immunohistology, and corneal topography were performed. RESULTS: In vivo confocal microscopy showed an irregular network of highly reflective structures representing activated keratocytes, which could be seen by light microscopy and characterized immunohistologically as myofibroblasts. Unstructured areas with increased reflectivity correlated with irregularly arranged collagen fibers and hyaline deposits in the nodulus. Epithelial cells in vivo appeared atypically shaped and elongated. These observations were consistent with decreased thickness of the epithelium over the nodules showed by histopathology. Treatment led to a dramatic reduction of hyperopia. Two months after surgery, uncorrected visual acuity (UCVA) in the right eye was 20/32 and 20/20 with a refraction of -0.75 -0.75/0 degrees. UCVA in the left eye was 20/40 and 20/20 with a refraction of +0.50 -1.75/165 degrees. Corneal topography showed regular astigmatism. CONCLUSION: In vivo confocal microscopy confirmed the clinicopathologic findings of Salzmann's nodular degeneration. Observations by in vivo confocal microscopy were consistent with the histopathologic descriptions of SND.

Corneal trephination with the femtosecond laser.

PURPOSE: To evaluate the feasibility and cut quality of corneal trephination in human donor corneal tissue with the femtosecond laser. METHODS: Twelve human corneoscleral discs were inserted in an artificial anterior chamber. After corneal thickness measurement and tonometry, the cornea was mounted on a femtosecond laser (FEMTEC; 20/10 Perfect Vision, Heidelberg, Germany) through a contact lens (patient interface). Trephination was performed with diameters of 7.0, 7.5, 8.0, and 8.5 mm in 3 corneas each. The corneal button was removed from the corneoscleral disc in 2 of the 3 corneas in each case. The cut was not manipulated in the remaining corneas to enable histologic detection of possible tissue bridges. The cut edges were macroscopically and light-microscopically examined for quality. RESULTS: Corneal buttons and corneoscleral discs could be separated by blunt dissection in all cases. Tissue bridges were more common in thicker edematous corneas than in thinner ones. Both the macro- and microscopic examination disclosed smooth rectilinear cut margins with a perpendicular cut edge. CONCLUSION: This feasibility study shows that the femtosecond laser enables sufficient trephination of human donor corneas.

Vitrification of posterior corneal lamellae.

Cryopreservation of corneas has not yet been established as a routine method. Unsatisfactory experimental results with conventional techniques prompted us to explore the possibilities of vitrification. The aim of the present study was to optimize the heat exchange between the corneal tissue and cooling medium by reducing the corneal tissue volume and using a suitable sample container. A further objective was to promote vitrification by developing a new device for rapid cooling to -140 degrees C, just below the vitrification temperature of the cryopreservation medium. Experiments were done using posterior lamellar discs from pig corneas with a diameter of 7.5 mm and a thickness of 250-350 microm. The volume of tissue to be vitrified was 88% lower with posterior corneal lamellae than with the previously used corneoscleral discs. A very thin-walled (0.05 mm), teflon-coated bag served as the sample container. Immersed in only 0.1 ml of the vitrification solution VS41a, the lamellae were cooled to a final storage temperature of -196 degrees C. After warming and organ-culturing for 24h, the endothelium was stained with trypan blue and alizarin red, to determine cell viability. Vitrification of corneal lamellae without apparent ice formation or cracking of the specimen was achieved. Despite the successful vitrification, only a maximum of 10% of the endothelial cells was vital after warming. Thus, the toxicity of the cryoprotective agents and the devitrification that occurred during the heating process require further optimization of the method.