Search

Abstract There is a critical need to generate environmentally relevant microplastics (MPs) and nanoplastics (NPs) to better investigate their behavior in laboratory settings. Environmental MPs are heterogenous in size and shape, unlike monodisperse and uniform microspheres commonly used as surrogates. Cryogenic grinding, or cryomilling, was successfully utilized to transform polystyrene (PS) bulk material into heterogenous micro and nano fragments. Fourier-Transform Infrared (FTIR) spectroscopy confirmed that this approach did not alter polymer surface chemistry. The number of milling cycles (time of milling) and frequency of grinding (intensity of milling) were varied to investigate the role cryomilling parameters had on generated MP characteristics. The resulting particle size distributions of cryomilled samples were measured and compared. Coulter Counter and Nanoparticle Tracking Analysis (NTA) were used to measure the particle size distributions at the micro and nanoparticle size ranges, respectively. Microspheres were used to determine what camera settings yielded more accurate sizing and to reduce bias in the NTA analysis. Increasing milling cycles generally increased the number of smaller particles. The evolution of the measured size distributions indicated that small nanosized fragments broke off from larger MPs during cryomilling, steadily eroding larger MP fragments. The number of milling cycles was observed to more consistently impact the size distributions of fragments compared to the frequency of milling. This study offers both analysis of the cryomilling process and recommendations for generating more realistic PS MP/NPs for examining environmental fate and effects.

Resolving natal populations for juvenile green turtles is challenging given their potential for extensive dispersal during the oceanic stage and ontogenetic shifts among nursery habitats. Mitochondrial DNA markers have elucidated patterns of connectivity between green turtle nesting populations (rookeries) and juvenile foraging aggregations. However, missing rookery baseline data and haplotype sharing among populations have often impeded inferences, including estimating origins of Gulf of Mexico juveniles. Here, we assessed genetic structure among seven foraging aggregations spanning southern Texas (TX) to southwestern Florida (SWFL), including Port Fourchon, Louisiana (LA); a surface-pelagic aggregation (SP) offshore of Louisiana and Florida; Santa Rosa Island, Florida (SRI); St. Joseph Bay, Florida (SJB); and the Big Bend region, Florida (BB). We estimated source contributions to aggregations with novel genetic data (excluding SP and BB) using a Bayesian many-to-one mixed stock analysis (MSA) approach. Haplotype frequencies for western (TX, LA, SP, SRI) and eastern (SJB, BB, SWFL) aggregations were significantly differentiated. The largest shift in haplotype frequencies between proximal nursery sites occurred between SRI and SJB, separated by only 150 km, highlighting the lack of a geographic yardstick for predicting genetic structure. In contrast to previous MSA results, there was no signal of Florida juveniles at any foraging site. Mexican contributions dominated in all aggregations, with strong connectivity between western Bay of Campeche (Tamaulipas/Veracruz) rookeries and western foraging aggregations. MSA indicated more diverse Mexican origins for eastern aggregations, with larger inputs from the eastern Bay of Campeche (Campeche/Yucatán), Campeche Bank, and Quintana Roo rookeries. These results demonstrate the significance of the Gulf of Mexico coast and offshore waters of the United States as important nursery habitat for green turtles of Mexican origin and highlight the need for international coordination for management of these populations.

Brennan Beal,1 Vanessa Shih,2 Joanna Campbell,2 David Veenstra,1 Beth Devine1 1The Comparative Health Outcomes, Policy and Economics (CHOICE) Institute, University of Washington, Seattle, WA, USA; 2Healthcare Economics and Outcomes Research, AbbVie, Chicago, IL, USACorrespondence: Brennan BealThe Comparative Health Outcomes, Policy and Economics (CHOICE) Institute, University of Washington, Suite H-375, Box 35760, 1959 NE Pacific Street, Seattle, WA, 98195-7630, USATel +1 704 742 0288Fax +1 206-543-3835Email brennanbeal@gmail.comPurpose: To estimate the healthcare resource utilization (HCRU) and costs for patients with normal tension glaucoma (NTG) as well as their payers across various levels of disease severity.Patients and Methods: Our study was a retrospective cohort study of 6330 US NTG patients. Patients were enrolled if they were 40 years or older and had two or more qualifying NTG diagnoses within the enrollment period, October 1st, 2015 to December 31st, 2017. Our analysis was carried out for two cohorts – those with unilateral disease and those with bilateral disease. Baseline demographic and clinical characteristics were assessed for a 12-month pre-index period. The follow up period was 12-months post-index. We employed generalized linear models to model HCRU and costs.Results: Patients with severe, bilateral disease, filled more than two additional prescriptions annually (2.5, p< 0.001, 95% CI [2.0, 3.1]) when compared to their mild counterparts and accounted for 111 (p< 0.001, 95% CI [83.5, 139.1]) extra days of supply of glaucoma medications. These patients face an adjusted $187 (p< 0.001, 95% CI [145, 229]) more out-of-pocket (OOP), and payers an additional $598 (p< 0.001, 95% CI [$370, $826]), than their counterparts with a mild diagnosis on an annual basis. Total annual payer costs, on average (SD) for those with severe bilateral NTG were $1175 ($2222).Conclusion: Our results suggest that patient and payer burden is significantly greater for those with severe disease compared to those with mild NTG. The excess burden is attributed to additional HCRU and the associated financial burden. Payers experienced a much larger financial burden from patients with severe disease compared to those with mild NTG. Approximately half of the cost differences can be attributed to additional prescription use.Keywords: healthcare resource utilization, normal tension glaucoma, low tension glaucoma, healthcare costs

Judy Munday,1– 4 Niall Higgins,1,2,5 Saira Mathew,1,2 Lizanne Dalgleish,1,2,5 Anthony S Batterbury,1,2,5 Luke Burgess,1,2,4 Jill Campbell,1,2,5 Lori J Delaney,1,2,6 Bronwyn R Griffin,1,2 James A Hughes,1,2,5 Jessica Ingleman,1,2 Samantha Keogh,1,2,5,7 Fiona Coyer1,2,5 1Centre for Healthcare Transformation, Faculty of Health, Queensland University of Technology (QUT), Brisbane, QLD, Australia; 2School of Nursing, Queensland University of Technology (QUT), Brisbane, QLD, Australia; 3 Department of Health and Nursing Science, University of Agder, Grimstad, 4879, Norway; 4Mater Research Institute-UQ, South Brisbane, QLD 4101, Australia; 5Royal Brisbane and Women’s Hospital, Herston, QLD 4029, Australia; 6Colleges of Health and Medicine, Australian National University, Acton, ACT 2601, Australia; 7Alliance for Vascular Access Teaching and Research, Griffith University, Brisbane, QLD, 4111, AustraliaCorrespondence: Judy MundaySchool of Nursing, Rm 529, Level 5, N Block, Victoria Park Road, Kelvin Grove, QLD 4059, AustraliaTel +61 3138 8209Email judy.munday@qut.edu.auPurpose: Nurses provide care at each phase of the complex, perioperative pathway and are well placed to identify areas of care requiring investigation in randomized controlled trials. Yet, currently, the scope of nurse-led randomized controlled trials conducted within the perioperative setting are unknown. This scoping review aims to identify areas of perioperative care in which nurse-led randomized controlled trials have been conducted, to identify issues impacting upon the quality of these trials and identify gaps for future investigation.Methods: This scoping review was conducted in reference to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews. Searches were conducted in PubMed, Embase, Cumulative Index for Nursing and Allied Health Literature and the Cochrane Central Register of Controlled Trials, with a date range of 2014– 19. Sources of unpublished literature included Open Grey, and ProQuest Dissertation and Theses, Clinical Trials.gov and the Australian and New Zealand Clinical Trials Registry. After title and abstract checking, full-text retrieval and data extraction, studies were appraised using the Joanna Briggs Institute Critical Appraisal Checklists for randomized controlled trials. Data were synthesized according to the main objectives. Key information was tabulated.Results: From the 86 included studies, key areas where nurses have led randomized controlled trials include patient or caregiver anxiety; postoperative pain relief; surgical site infection prevention: patient and caregiver knowledge; perioperative hypothermia prevention; postoperative nausea and vomiting; in addition to other diverse outcomes. Issues impacting upon quality (including poorly reported randomization), and gaps for future investigation (including a focus on vulnerable populations), are evident.Conclusion: Nurse-led randomized controlled trials in the perioperative setting have focused on key areas of perioperative care. Yet, opportunities exist for nurses to lead experimental research in other perioperative priority areas and within different populations that have been neglected, such as in the population of older adults undergoing surgery.Keywords: perioperative, nursing, randomized controlled trial, scoping review

Population assessments conducted at reproductive sites of migratory species necessitate understanding the foraging-area origins of breeding individuals. Without this information, efforts to contextualize changes in breeding populations and develop effective management strategies are compromised. We used stable isotope analysis of tissue samples collected from loggerhead sea turtles (Caretta caretta) nesting at seven sites in the Northern Recovery Unit (NRU) of the eastern United States (North Carolina, South Carolina and Georgia) to assign females to three separate foraging areas in the Northwest Atlantic Ocean (NWA). We found that the majority of the females at NRU nesting sites (84.4%) use more northern foraging areas in the Mid-Atlantic Bight, while fewer females use more proximate foraging areas in the South Atlantic Bight (13.4%) and more southerly foraging areas in the Subtropical Northwest Atlantic (2.2%). We did not find significant latitudinal or temporal trends in the proportions of NRU females originating from different foraging areas. Combining these findings with previous data from stable isotope and satellite tracking studies across NWA nesting sites showed that variation in the proportion of adult loggerheads originating from different foraging areas is primarily related differences between recovery units: individuals in the NRU primarily use the Mid-Atlantic Bight foraging area, while individuals from the three Florida recovery units primarily use the Subtropical Northwest Atlantic and Eastern Gulf of Mexico foraging areas. Because each foraging area is associated with its own distinct ecological characteristics, environmental fluctuations and anthropogenic threats that affect the abundance and productivity of individuals at nesting sites, this information is critical for accurately evaluating population trends and developing effective region-specific management strategies.

PURPOSE: The murine oxygen-induced retinopathy (OIR) model is one of the most widely used animal models of ischemic retinopathy, mimicking hallmark pathophysiology of initial vaso-obliteration (VO) resulting in ischemia that drives neovascularization (NV). In addition to NV and VO, human ischemic retinopathies, including retinopathy of prematurity (ROP), are characterized by increased vascular tortuosity. Vascular tortuosity is an indicator of disease severity, need to treat, and treatment response in ROP. Current literature investigating novel therapeutics in the OIR model often report their effects on NV and VO, and measurements of vascular tortuosity are less commonly performed. No standardized quantification of vascular tortuosity exists to date despite this metrics relevance to human disease. This proof-of-concept study aimed to apply a previously published semi-automated computer-based image analysis approach (iROP-Assist) to develop a new tool to quantify vascular tortuosity in mouse models. DESIGN: Experimental study. SUBJECTS: C57BL/6J mice subjected to the OIR model. METHODS: In a pilot study, vasculature was manually segmented on flat-mount images of OIR and normoxic (NOX) mice retinas and segmentations were analyzed with iROP-Assist to quantify vascular tortuosity metrics. In a large cohort of age-matched (postnatal day 12 [P12], P17, P25) NOX and OIR mice retinas, NV, VO, and vascular tortuosity were quantified and compared. In a third experiment, vascular tortuosity in OIR mice retinas was quantified on P17 following intravitreal injection with anti-VEGF (aflibercept) or Immunoglobulin G isotype control on P12. MAIN OUTCOME MEASURES: Vascular tortuosity. RESULTS: Cumulative tortuosity index was the best metric produced by iROP-Assist for discriminating between OIR mice and NOX controls. Increased vascular tortuosity correlated with disease activity in OIR. Treatment of OIR mice with aflibercept rescued vascular tortuosity. CONCLUSIONS: Vascular tortuosity is a quantifiable feature of the OIR model that correlates with disease severity and may be quickly and accurately quantified using the iROP-Assist algorithm. FINANCIAL DISCLOSURES: Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

We conducted a genome-wide association study (GWAS) in a multiethnic cohort of 920 at-risk infants for retinopathy of prematurity (ROP), a major cause of childhood blindness, identifying 1 locus at genome-wide significance level (p

OBJECTIVE: To develop a generative adversarial network (GAN) to segment major blood vessels from retinal flat-mount images from oxygen-induced retinopathy (OIR) and demonstrate the utility of these GAN-generated vessel segmentations in quantifying vascular tortuosity. DESIGN: Development and validation of GAN. SUBJECTS: Three datasets containing 1084, 50, and 20 flat-mount mice retina images with various stains used and ages at sacrifice acquired from previously published manuscripts. METHODS: Four graders manually segmented major blood vessels from flat-mount images of retinas from OIR mice. Pix2Pix, a high-resolution GAN, was trained on 984 pairs of raw flat-mount images and manual vessel segmentations and then tested on 100 and 50 image pairs from a held-out and external test set, respectively. GAN-generated and manual vessel segmentations were then used as an input into a previously published algorithm (iROP-Assist) to generate a vascular cumulative tortuosity index (CTI) for 20 image pairs containing mouse eyes treated with aflibercept versus control. MAIN OUTCOME MEASURES: Mean dice coefficients were used to compare segmentation accuracy between the GAN-generated and manually annotated segmentation maps. For the image pairs treated with aflibercept versus control, mean CTIs were also calculated for both GAN-generated and manual vessel maps. Statistical significance was evaluated using Wilcoxon signed-rank tests (P ≤ 0.05 threshold for significance). RESULTS: The dice coefficient for the GAN-generated versus manual vessel segmentations was 0.75 ± 0.27 and 0.77 ± 0.17 for the held-out test set and external test set, respectively. The mean CTI generated from the GAN-generated and manual vessel segmentations was 1.12 ± 0.07 versus 1.03 ± 0.02 (P = 0.003) and 1.06 ± 0.04 versus 1.01 ± 0.01 (P < 0.001), respectively, for eyes treated with aflibercept versus control, demonstrating that vascular tortuosity was rescued by aflibercept when quantified by GAN-generated and manual vessel segmentations. CONCLUSIONS: GANs can be used to accurately generate vessel map segmentations from flat-mount images. These vessel maps may be used to evaluate novel metrics of vascular tortuosity in OIR, such as CTI, and have the potential to accelerate research in treatments for ischemic retinopathies. FINANCIAL DISCLOSURES: The author(s) have no proprietary or commercial interest in any materials discussed in this article.

The Canada warbler (Cardellina canadensis) is a migratory neotropical warbler species of conservation concern. The 16,807 bp genome was sequenced and contains a 1224 bp control region, 37 genes, and has a nucleotide composition of 30.6% A, 23.7% T, 14.1% G, and 31.6% C. We constructed a phylogenetic tree of species in the order Passeriformes, which placed C. canadensis closest to species of Icteridae and Emberizidae. As the first complete mitogenome of any species in the family Parulidae, this mitogenome sequence provides a resource for future genetic and phylogenetic research involving C. canadensis and other Parulidae.

BACKGROUND: Antimicrobial resistance (AMR) is undermining modern medicine, a problem compounded by bacterial adaptation to antibiotic pressures. Phages are viruses that infect bacteria. Their diversity and evolvability offer the prospect of their use as a therapeutic solution. Reported are outcomes of customized phage therapy for patients with difficult-to-treat antimicrobial resistant infections. METHODS: We retrospectively assessed 12 cases of customized phage therapy from a phage production center. Phages were screened, purified, sequenced, characterized, and Food and Drug Administration-approved via the IND (investigational new drug) compassionate-care route. Outcomes were assessed as favorable or unfavorable by microbiologic and clinical standards. Infections were device-related or systemic. Other experiences such as time to treatment, antibiotic synergy, and immune responses were recorded. RESULTS: Fifty requests for phage therapy were received. Customized phages were generated for 12 patients. After treatment, 42% (5/12) of cases showed bacterial eradication and 58% (7/12) showed clinical improvement, with two-thirds of all cases (66%) showing favorable responses. No major adverse reactions were observed. Antibiotic-phage synergy in vitro was observed in most cases. Immunological neutralization of phages was reported in 5 cases. Several cases were complicated by secondary infections. Complete characterization of the phages (morphology, genomics, and activity) and their production (methods, sterility, and endotoxin tests) are reported. CONCLUSIONS: Customized phage production and therapy was safe and yielded favorable clinical or microbiological outcomes in two-thirds of cases. A center or pipeline dedicated to tailoring the phages against a patients specific AMR bacterial infection may be a viable option where standard treatment has failed.

Szilárd Kiss,1 Ying Liu,2 Joseph Brown,3 Nancy M Holekamp,4,5 Arghavan Almony,6 Joanna Campbell,2 Jonathan W Kowalski2 1Weill Cornell Medical College, New York, NY; 2Allergan, Inc., Irvine, CA; 3IMS Health, Woodland Hills, CA; 4Pepose Vision Institute, Chesterfield, MO; 5Washington University School of Medicine, St Louis, MO; 6Carolina Eye Associates, Southern Pines, NC, USA Purpose: To examine the utilization of bevacizumab and ranibizumab and disease monitoring in patients with branch or central retinal vein occlusion (BRVO/CRVO) or diabetic macular edema (DME) in clinical practice.Patients and methods: This retrospective claims analysis included newly diagnosed patients with one or more bevacizumab or ranibizumab injections. Bevacizumab or ranibizumab utilization was assessed by year of first injection: 2008–2010 cohorts (12-month follow-up), January to June 2011 cohort (6-month follow-up). The main outcome measures were mean annual numbers of injections, ophthalmologist visits and optical coherence tomography examinations, and proportion of patients with additional laser or intravitreal triamcinolone (IVTA) use.Results: A total of 885 BRVO, 611 CRVO, and 2,733 DME patients treated with bevacizumab were included, with too few ranibizumab-treated patients for meaningful analysis. Across the 2008, 2009, and 2010 cohorts, mean annual numbers of bevacizumab injections increased, but remained low (BRVO 2.5, 3.1, 3.3; CRVO 3.1, 3.1, 3.5; and DME 2.2, 2.5, 3.6, respectively); mean ophthalmologist visits ranged between 4.4 and 6.5, and mean optical coherence tomography examinations ranged between 3.1 and 3.9 across all conditions. A total of 42.0% of BRVO, 16.5% of CRVO, and 57.7% of DME patients received additional laser or IVTA therapy. The number of bevacizumab injections was positively associated with laser use in BRVO (3.3 versus 2.9, P

Background: Pharmacist managed warfarin clinics can improve the anticoagulation status of non-valvular patients. The first of such services was implemented at the Cornwall Regional Hospital in Jamaica in 2013. Objectives: To assess the anticoagulation control of patients on warfarin therapy over six months in the pharmacist managed warfarin clinic at Cornwall Regional Hospital. Methods: Retrospective docket review for the period January 2014 to December 2016 was done to include data of patients attending routine clinic appointments for at least six months. Age, gender, date of visit, indication for warfarin therapy, warfarin dose and International Normalized Ratio readings were extracted. Percentage time spent in therapeutic range (TTR) was calculated by month for six months using the Rosendaal linear interpolation method. Patient anticoagulation status was categorized as poor (TTR40% and sustaining these values over six months. The services should therefore be encouraged.

The effects of predicted declines and potential loss of individual populations on species-level genetic diversity is unclear. A number of taxa, including the Canada Warbler (Cardellina canadensis), share wide-ranging geographic distributions in North American boreal forests with trailing-edge populations extending into the southern Appalachian Mountains. Trailing-edge populations in the southern portion of a species' ranges often harbor high levels of genetic diversity and unique genetic variants, and may be at risk of extinction from climate change. Climate change and other anthropogenic factors are causing declines in the Canada Warbler's southern trailing-edge populations, and with no genetic studies to date, the effect on species-level genetic diversity is uncertain. Species-specific microsatellite markers for the Canada Warbler were developed and validated using samples from three populations, including a southern trailing-edge population, to investigate their utility for intraspecific population studies. Eight of the microsatellite markers were informative for assessing genetic diversity and preliminary analysis suggests that they have potential for characterizing intraspecific neutral genetic diversity and structure among Canada Warbler populations.

PurposeTo assess changes in retinopathy of prematurity (ROP) diagnosis in single and serial retinal images.DesignCohort study.ParticipantsCases of ROP recruited from the Imaging and Informatics in Retinopathy of Prematurity (i-ROP) consortium evaluated by 7 graders.MethodsSeven ophthalmologists reviewed both single and 3 consecutive serial retinal images from 15 cases with ROP, and severity was assigned as plus, preplus, or none. Imaging data were acquired during routine ROP screening from 2011 to 2015, and a reference standard diagnosis was established for each image. A secondary analysis was performed using the i-ROP deep learning system to assign a vascular severity score (VSS) to each image, ranging from 1 to 9, with 9 being the most severe disease. This score has been previously demonstrated to correlate with the International Classification of ROP. Mean plus disease severity was calculated by averaging 14 labels per image in serial and single images to decrease noise.Main outcome measuresGrading severity of ROP as defined by plus, preplus, or no ROP.ResultsAssessment of serial retinal images changed the grading severity for > 50% of the graders, although there was wide variability. Cohen's kappa ranged from 0.29 to 1.0, which showed a wide range of agreement from slight to perfect by each grader. Changes in the grading of serial retinal images were noted more commonly in cases of preplus disease. The mean severity in cases with a diagnosis of plus disease and no disease did not change between single and serial images. The ROP VSS demonstrated good correlation with the range of expert classifications of plus disease and overall agreement with the mode class (P = 0.001). The VSS correlated with mean plus disease severity by expert diagnosis (correlation coefficient, 0.89). The more aggressive graders tended to be influenced by serial images to increase the severity of their grading. The VSS also demonstrated agreement with disease progression across serial images, which progressed to preplus and plus disease.ConclusionsClinicians demonstrated variability in ROP diagnosis when presented with both single and serial images. The use of deep learning as a quantitative assessment of plus disease has the potential to standardize ROP diagnosis and treatment.

In order to provide information to better inform management decisions and direct further research, vessel-based visual transects, snorkel transects, and in-water capture techniques were used to characterize hawksbill sea turtles in the shallow marine habitats of a Marine Protected Area (MPA), the Key West National Wildlife Refuge in the Florida Keys. Hawksbills were found in hardbottom and seagrass dominated habitats throughout the Refuge, and on man-made rubble structures in the Northwest Channel near Cottrell Key. Hawksbills captured (N = 82) were exclusively juveniles and subadults with a straight standard carapace length (SSCL) ranging from 21.4 to 69.0cm with a mean of 44.1 cm (SD = 10.8). Somatic growth rates were calculated from 15 recaptured turtles with periods at large ranging from 51 to 1188 days. Mean SSCL growth rate was 7.7 cm/year (SD = 4.6). Juvenile hawksbills (

Previous genetic studies have demonstrated that natal homing shapes the stock structure of marine turtle nesting populations. However, widespread sharing of common haplotypes based on short segments of the mitochondrial control region often limits resolution of the demographic connectivity of populations. Recent studies employing longer control region sequences to resolve haplotype sharing have focused on regional assessments of genetic structure and phylogeography. Here we synthesize available control region sequences for loggerhead turtles from the Mediterranean Sea, Atlantic, and western Indian Ocean basins. These data represent six of the nine globally significant regional management units (RMUs) for the species and include novel sequence data from Brazil, Cape Verde, South Africa and Oman. Genetic tests of differentiation among 42 rookeries represented by short sequences (380 bp haplotypes from 3,486 samples) and 40 rookeries represented by long sequences (∼800 bp haplotypes from 3,434 samples) supported the distinction of the six RMUs analyzed as well as recognition of at least 18 demographically independent management units (MUs) with respect to female natal homing. A total of 59 haplotypes were resolved. These haplotypes belonged to two highly divergent global lineages, with haplogroup I represented primarily by CC-A1, CC-A4, and CC-A11 variants and haplogroup II represented by CC-A2 and derived variants. Geographic distribution patterns of haplogroup II haplotypes and the nested position of CC-A11.6 from Oman among the Atlantic haplotypes invoke recent colonization of the Indian Ocean from the Atlantic for both global lineages. The haplotypes we confirmed for western Indian Ocean RMUs allow reinterpretation of previous mixed stock analysis and further suggest that contemporary migratory connectivity between the Indian and Atlantic Oceans occurs on a broader scale than previously hypothesized. This study represents a valuable model for conducting comprehensive international cooperative data management and research in marine ecology.

PurposeDevelop computer vision models for image-based differentiation of bacterial and fungal corneal ulcers and compare their performance against human experts.DesignCross-sectional comparison of diagnostic performance.ParticipantsPatients with acute, culture-proven bacterial or fungal keratitis from 4 centers in South India.MethodsFive convolutional neural networks (CNNs) were trained using images from handheld cameras collected from patients with culture-proven corneal ulcers in South India recruited as part of clinical trials conducted between 2006 and 2015. Their performance was evaluated on 2 hold-out test sets (1 single center and 1 multicenter) from South India. Twelve local expert cornea specialists performed remote interpretation of the images in the multicenter test set to enable direct comparison against CNN performance.Main outcome measuresArea under the receiver operating characteristic curve (AUC) individually and for each group collectively (i.e., CNN ensemble and human ensemble).ResultsThe best-performing CNN architecture was MobileNet, which attained an AUC of 0.86 on the single-center test set (other CNNs range, 0.68-0.84) and 0.83 on the multicenter test set (other CNNs range, 0.75-0.83). Expert human AUCs on the multicenter test set ranged from 0.42 to 0.79. The CNN ensemble achieved a statistically significantly higher AUC (0.84) than the human ensemble (0.76; P < 0.01). CNNs showed relatively higher accuracy for fungal (81%) versus bacterial (75%) ulcers, whereas humans showed relatively higher accuracy for bacterial (88%) versus fungal (56%) ulcers. An ensemble of the best-performing CNN and best-performing human achieved the highest AUC of 0.87, although this was not statistically significantly higher than the best CNN (0.83; P = 0.17) or best human (0.79; P = 0.09).ConclusionsComputer vision models achieved superhuman performance in identifying the underlying infectious cause of corneal ulcers compared with cornea specialists. The best-performing model, MobileNet, attained an AUC of 0.83 to 0.86 without any additional clinical or historical information. These findings suggest the potential for future implementation of these models to enable earlier directed antimicrobial therapy in the management of infectious keratitis, which may improve visual outcomes. Additional studies are ongoing to incorporate clinical history and expert opinion into predictive models.

We report high-precision measurements of the deeply virtual Compton scattering (DVCS) cross section at high values of the Bjorken variable x_{B}. DVCS is sensitive to the generalized parton distributions of the nucleon, which provide a three-dimensional description of its internal constituents. Using the exact analytic expression of the DVCS cross section for all possible polarization states of the initial and final electron and nucleon, and final state photon, we present the first experimental extraction of all four helicity-conserving Compton form factors (CFFs) of the nucleon as a function of x_{B}, while systematically including helicity flip amplitudes. In particular, the high accuracy of the present data demonstrates sensitivity to some very poorly known CFFs.

We present new precision measurements of the elastic electron-proton scattering cross section for momentum transfer (Q^{2}) up to 15.75  (GeV/c)^{2}. Combined with existing data, these provide an improved extraction of the proton magnetic form factor at high Q^{2} and double the range over which a longitudinal or transverse separation of the cross section can be performed. The difference between our results and polarization data agrees with that observed at lower Q^{2} and attributed to hard two-photon exchange (TPE) effects, extending to 8  (GeV/c)^{2} the range of Q^{2} for which a discrepancy is established at >95% confidence. We use the discrepancy to quantify the size of TPE contributions needed to explain the cross section at high Q^{2}.