Choroidalyzer: An Open-Source, End-to-End Pipeline for Choroidal Analysis in Optical Coherence Tomography.

Purpose: To develop Choroidalyzer, an open-source, end-to-end pipeline for segmenting the choroid region, vessels, and fovea, and deriving choroidal thickness, area, and vascular index. Methods: We used 5600 OCT B-scans (233 subjects, six systemic disease cohorts, three device types, two manufacturers). To generate region and vessel ground-truths, we used state-of-the-art automatic methods following manual correction of inaccurate segmentations, with foveal positions manually annotated. We trained a U-Net deep learning model to detect the region, vessels, and fovea to calculate choroid thickness, area, and vascular index in a fovea-centered region of interest. We analyzed segmentation agreement (AUC, Dice) and choroid metrics agreement (Pearson, Spearman, mean absolute error [MAE]) in internal and external test sets. We compared Choroidalyzer to two manual graders on a small subset of external test images and examined cases of high error. Results: Choroidalyzer took 0.299 seconds per image on a standard laptop and achieved excellent region (Dice: internal 0.9789, external 0.9749), very good vessel segmentation performance (Dice: internal 0.8817, external 0.8703), and excellent fovea location prediction (MAE: internal 3.9 pixels, external 3.4 pixels). For thickness, area, and vascular index, Pearson correlations were 0.9754, 0.9815, and 0.8285 (internal)/0.9831, 0.9779, 0.7948 (external), respectively (all P < 0.0001). Choroidalyzer's agreement with graders was comparable to the intergrader agreement across all metrics. Conclusions: Choroidalyzer is an open-source, end-to-end pipeline that accurately segments the choroid and reliably extracts thickness, area, and vascular index. Especially choroidal vessel segmentation is a difficult and subjective task, and fully automatic methods like Choroidalyzer could provide objectivity and standardization.

PallorMetrics: Software for Automatically Quantifying Optic Disc Pallor in Fundus Photographs, and Associations With Peripapillary RNFL Thickness.

Purpose: We sough to develop an automatic method of quantifying optic disc pallor in fundus photographs and determine associations with peripapillary retinal nerve fiber layer (pRNFL) thickness. Methods: We used deep learning to segment the optic disc, fovea, and vessels in fundus photographs, and measured pallor. We assessed the relationship between pallor and pRNFL thickness derived from optical coherence tomography scans in 118 participants. Separately, we used images diagnosed by clinical inspection as pale (n = 45) and assessed how measurements compared with healthy controls (n = 46). We also developed automatic rejection thresholds and tested the software for robustness to camera type, image format, and resolution. Results: We developed software that automatically quantified disc pallor across several zones in fundus photographs. Pallor was associated with pRNFL thickness globally (ß = -9.81; standard error [SE] = 3.16; P < 0.05), in the temporal inferior zone (ß = -29.78; SE = 8.32; P < 0.01), with the nasal/temporal ratio (ß = 0.88; SE = 0.34; P < 0.05), and in the whole disc (ß = -8.22; SE = 2.92; P < 0.05). Furthermore, pallor was significantly higher in the patient group. Last, we demonstrate the analysis to be robust to camera type, image format, and resolution. Conclusions: We developed software that automatically locates and quantifies disc pallor in fundus photographs and found associations between pallor measurements and pRNFL thickness. Translational Relevance: We think our method will be useful for the identification, monitoring, and progression of diseases characterized by disc pallor and optic atrophy, including glaucoma, compression, and potentially in neurodegenerative disorders.

Evaluation of an Automated Choroid Segmentation Algorithm in a Longitudinal Kidney Donor and Recipient Cohort.

Purpose: To evaluate the performance of an automated choroid segmentation algorithm in optical coherence tomography (OCT) data using a longitudinal kidney donor and recipient cohort. Methods: We assessed 22 donors and 23 patients requiring renal transplantation over up to 1 year posttransplant. We measured choroidal thickness (CT) and area and compared our automated CT measurements to manual ones at the same locations. We estimated associations between choroidal measurements and markers of renal function (estimated glomerular filtration rate [eGFR], serum creatinine, and urea) using correlation and linear mixed-effects (LME) modeling. Results: There was good agreement between manual and automated CT. Automated measures were more precise because of smaller measurement error over time. External adjudication of major discrepancies was in favor of automated measures. Significant differences were observed in the choroid pre- and posttransplant in both cohorts, and LME modeling revealed significant linear associations observed between choroidal measures and renal function in recipients. Significant associations were mostly stronger with automated CT (eGFR, P < 0.001; creatinine, P = 0.004; urea, P = 0.04) compared to manual CT (eGFR, P = 0.002; creatinine, P = 0.01; urea, P = 0.03). Conclusions: Our automated approach has greater precision than human-performed manual measurements, which may explain stronger associations with renal function compared to manual measurements. To improve detection of meaningful associations with clinical endpoints in longitudinal studies of OCT, reducing measurement error should be a priority, and automated measurements help achieve this. Translational Relevance: We introduce a novel choroid segmentation algorithm that can replace manual grading for studying the choroid in renal disease and other clinical conditions.

An Open-Source Deep Learning Algorithm for Efficient and Fully Automatic Analysis of the Choroid in Optical Coherence Tomography.

Purpose: To develop an open-source, fully automatic deep learning algorithm, DeepGPET, for choroid region segmentation in optical coherence tomography (OCT) data. Methods: We used a dataset of 715 OCT B-scans (82 subjects, 115 eyes) from three clinical studies related to systemic disease. Ground-truth segmentations were generated using a clinically validated, semiautomatic choroid segmentation method, Gaussian Process Edge Tracing (GPET). We finetuned a U-Net with the MobileNetV3 backbone pretrained on ImageNet. Standard segmentation agreement metrics, as well as derived measures of choroidal thickness and area, were used to evaluate DeepGPET, alongside qualitative evaluation from a clinical ophthalmologist. Results: DeepGPET achieved excellent agreement with GPET on data from three clinical studies (AUC = 0.9994, Dice = 0.9664; Pearson correlation = 0.8908 for choroidal thickness and 0.9082 for choroidal area), while reducing the mean processing time per image on a standard laptop CPU from 34.49 ± 15.09 seconds using GPET to 1.25 ± 0.10 seconds using DeepGPET. Both methods performed similarly according to a clinical ophthalmologist who qualitatively judged a subset of segmentations by GPET and DeepGPET, based on smoothness and accuracy of segmentations. Conclusions: DeepGPET, a fully automatic, open-source algorithm for choroidal segmentation, will enable researchers to efficiently extract choroidal measurements, even for large datasets. As no manual interventions are required, DeepGPET is less subjective than semiautomatic methods and could be deployed in clinical practice without requiring a trained operator. Translational Relevance: DeepGPET addresses the lack of open-source, fully automatic, and clinically relevant choroid segmentation algorithms, and its subsequent public release will facilitate future choroidal research in both ophthalmology and wider systemic health.

Guardians and research staff experiences and views about the consent process in hospital-based paediatric research studies in urban Malawi: A qualitative study.

BACKGROUND: Obtaining consent has become a standard way of respecting the patient's rights and autonomy in clinical research. Ethical guidelines recommend that the child's parent/s or authorised legal guardian provides informed consent for their child's participation. However, obtaining informed consent in paediatric research is challenging. Parents become vulnerable because of stress related to their child's illness. Understanding the views held by guardians and researchers about the consent process in Malawi, where there are limitations in health care access and research literacy will assist in developing appropriate consent guidelines. METHODS: We conducted 20 in-depth interviews with guardians of children and research staff who had participated in paediatric clinical trial and observational studies in acute and non-acute settings in the Southern Region of Malawi. Interviews were audio-recorded, transcribed verbatim, and thematically analysed. Interviews were compared across studies and settings to identify differences and similarities in participants' views about informed consent processes. Data analysis was facilitated by NVIVO 11 software. RESULTS: All participants across study types and settings reported that they associated participating in research with therapeutic benefits. Substantial differences were noted in the decision-making process across study settings. Guardians from acute studies felt that the role of their spouses was neglected during consenting, while staff reported that they had problems obtaining consent from guardians when their partners were not present. Across all study types and settings, research staff reported that they emphasised the benefits more than the risks of the study to participants, due to pressure to recruit. Participants from non-acute settings were more likely to recall information shared during the consent process than participants in the acute setting. CONCLUSION: The health care context, culture and research process influenced participants' understanding of study information across study types and settings. We advise research managers or principal investigators to define minimum requirements that would not compromise the consent process and conduct study specific training for staff. The use of one size fits all consent process may not be ideal. More guidance is needed on how these differences can be incorporated during the consent process to improve understanding and delivery of consent. Trial registration Not applicable.

How Does Blood-Retinal Barrier Breakdown Relate to Death and Disability in Pediatric Cerebral Malaria?

BACKGROUND: In cerebral malaria, the retina can be used to understand disease pathogenesis. The mechanisms linking sequestration, brain swelling, and death remain poorly understood. We hypothesized that retinal vascular leakage would be associated with brain swelling. METHODS: We used retinal angiography to study blood-retinal barrier integrity. We analyzed retinal leakage, histopathology, brain magnatic resonance imaging (MRI), and associations with death and neurological disability in prospective cohorts of Malawian children with cerebral malaria. RESULTS: Three types of retinal leakage were seen: large focal leak (LFL), punctate leak (PL), and vessel leak. The LFL and PL were associated with death (odds ratio [OR] = 13.20, 95% confidence interval [CI] = 5.21-33.78 and OR = 8.58, 95% CI = 2.56-29.08, respectively) and brain swelling (P < .05). Vessel leak and macular nonperfusion were associated with neurological disability (OR = 3.71, 95% CI = 1.26-11.02 and OR = 9.06, 95% CI = 1.79-45.90). Large focal leak was observed as an evolving retinal hemorrhage. A core of fibrinogen and monocytes was found in 39 (93%) white-centered hemorrhages. CONCLUSIONS: Blood-retina barrier breakdown occurs in 3 patterns in cerebral malaria. Associations between LFL, brain swelling, and death suggest that the rapid accumulation of cerebral hemorrhages, with accompanying fluid egress, may cause fatal brain swelling. Vessel leak, from barrier dysfunction, and nonperfusion were not associated with severe brain swelling but with neurological deficits, suggesting hypoxic injury in survivors.

Graph-Based Region and Boundary Aggregation for Biomedical Image Segmentation.

Segmentation is a fundamental task in biomedical image analysis. Unlike the existing region-based dense pixel classification methods or boundary-based polygon regression methods, we build a novel graph neural network (GNN) based deep learning framework with multiple graph reasoning modules to explicitly leverage both region and boundary features in an end-to-end manner. The mechanism extracts discriminative region and boundary features, referred to as initialized region and boundary node embeddings, using a proposed Attention Enhancement Module (AEM). The weighted links between cross-domain nodes (region and boundary feature domains) in each graph are defined in a data-dependent way, which retains both global and local cross-node relationships. The iterative message aggregation and node update mechanism can enhance the interaction between each graph reasoning module's global semantic information and local spatial characteristics. Our model, in particular, is capable of concurrently addressing region and boundary feature reasoning and aggregation at several different feature levels due to the proposed multi-level feature node embeddings in different parallel graph reasoning modules. Experiments on two types of challenging datasets demonstrate that our method outperforms state-of-the-art approaches for segmentation of polyps in colonoscopy images and of the optic disc and optic cup in colour fundus images. The trained models will be made available at: https://github.com/smallmax00/Graph_Region_Boudnary.

Correction: Accurate, fast, data efficient and interpretable glaucoma diagnosis with automated spatial analysis of the whole cup to disc profile.

[This corrects the article DOI: 10.1371/journal.pone.0209409.].

Accurate, fast, data efficient and interpretable glaucoma diagnosis with automated spatial analysis of the whole cup to disc profile.

BACKGROUND: Glaucoma is the leading cause of irreversible blindness worldwide. It is a heterogeneous group of conditions with a common optic neuropathy and associated loss of peripheral vision. Both over and under-diagnosis carry high costs in terms of healthcare spending and preventable blindness. The characteristic clinical feature of glaucoma is asymmetrical optic nerve rim narrowing, which is difficult for humans to quantify reliably. Strategies to improve and automate optic disc assessment are therefore needed to prevent sight loss. METHODS: We developed a novel glaucoma detection algorithm that segments and analyses colour photographs to quantify optic nerve rim consistency around the whole disc at 15-degree intervals. This provides a profile of the cup/disc ratio, in contrast to the vertical cup/disc ratio in common use. We introduce a spatial probabilistic model, to account for the optic nerve shape, we then use this model to derive a disc deformation index and a decision rule for glaucoma. We tested our algorithm on two separate image datasets (ORIGA and RIM-ONE). RESULTS: The spatial algorithm accurately distinguished glaucomatous and healthy discs on internal and external validation (AUROC 99.6% and 91.0% respectively). It achieves this using a dataset 100-times smaller than that required for deep learning algorithms, is flexible to the type of cup and disc segmentation (automated or semi-automated), utilises images with missing data, and is correlated with the disc size (p = 0.02) and the rim-to-disc at the narrowest rim (p<0.001, in external validation). DISCUSSION: The spatial probabilistic algorithm is highly accurate, highly data efficient and it extends to any imaging hardware in which the boundaries of cup and disc can be segmented, thus making the algorithm particularly applicable to research into disease mechanisms, and also glaucoma screening in low resource settings.

Spatial statistical modelling of capillary non-perfusion in the retina.

Manual grading of lesions in retinal images is relevant to clinical management and clinical trials, but it is time-consuming and expensive. Furthermore, it collects only limited information - such as lesion size or frequency. The spatial distribution of lesions is ignored, even though it may contribute to the overall clinical assessment of disease severity, and correspond to microvascular and physiological topography. Capillary non-perfusion (CNP) lesions are central to the pathogenesis of major causes of vision loss. Here we propose a novel method to analyse CNP using spatial statistical modelling. This quantifies the percentage of CNP-pixels in each of 48 sectors and then characterises the spatial distribution with goniometric functions. We applied our spatial approach to a set of images from patients with malarial retinopathy, and found it compares favourably with the raw percentage of CNP-pixels and also with manual grading. Furthermore, we were able to quantify a biological characteristic of macular CNP in malaria that had previously only been described subjectively: clustering at the temporal raphe. Microvascular location is likely to be biologically relevant to many diseases, and so our spatial approach may be applicable to a diverse range of pathological features in the retina and other organs.

Linking EPCR-Binding PfEMP1 to Brain Swelling in Pediatric Cerebral Malaria.

Brain swelling is a major predictor of mortality in pediatric cerebral malaria (CM). However, the mechanisms leading to swelling remain poorly defined. Here, we combined neuroimaging, parasite transcript profiling, and laboratory blood profiles to develop machine-learning models of malarial retinopathy and brain swelling. We found that parasite var transcripts encoding endothelial protein C receptor (EPCR)-binding domains, in combination with high parasite biomass and low platelet levels, are strong indicators of CM cases with malarial retinopathy. Swelling cases presented low platelet levels and increased transcript abundance of parasite PfEMP1 DC8 and group A EPCR-binding domains. Remarkably, the dominant transcript in 50% of swelling cases encoded PfEMP1 group A CIDRα1.7 domains. Furthermore, a recombinant CIDRα1.7 domain from a pediatric CM brain autopsy inhibited the barrier-protective properties of EPCR in human brain endothelial cells in vitro. Together, these findings suggest a detrimental role for EPCR-binding CIDRα1 domains in brain swelling.

Decreased Rate of Plasma Arginine Appearance in Murine Malaria May Explain Hypoargininemia in Children With Cerebral Malaria.

BACKGROUND: Plasmodium infection depletes arginine, the substrate for nitric oxide synthesis, and impairs endothelium-dependent vasodilation. Increased conversion of arginine to ornithine by parasites or host arginase is a proposed mechanism of arginine depletion. METHODS: We used high-performance liquid chromatography to measure plasma arginine, ornithine, and citrulline levels in Malawian children with cerebral malaria and in mice infected with Plasmodium berghei ANKA with or without the arginase gene. Heavy isotope-labeled tracers measured by quadrupole time-of-flight liquid chromatography-mass spectrometry were used to quantify the in vivo rate of appearance and interconversion of plasma arginine, ornithine, and citrulline in infected mice. RESULTS: Children with cerebral malaria and P. berghei-infected mice demonstrated depletion of plasma arginine, ornithine, and citrulline. Knock out of Plasmodium arginase did not alter arginine depletion in infected mice. Metabolic tracer analysis demonstrated that plasma arginase flux was unchanged by P. berghei infection. Instead, infected mice exhibited decreased rates of plasma arginine, ornithine, and citrulline appearance and decreased conversion of plasma citrulline to arginine. Notably, plasma arginine use by nitric oxide synthase was decreased in infected mice. CONCLUSIONS: Simultaneous arginine and ornithine depletion in malaria parasite-infected children cannot be fully explained by plasma arginase activity. Our mouse model studies suggest that plasma arginine depletion is driven primarily by a decreased rate of appearance.

Longitudinal Visuomotor Development in a Malaria Endemic Area: Cerebral Malaria and Beyond.

Paediatric cerebral malaria is the most serious complication of Plasmodium falciparum infection. While the majority recover, long-term cognitive impairment has been highlighted as a significant and neglected problem. Persistent or serious deficits in processes such as attention or behavioural inhibition should be manifest in changes to performance on oculomotor tasks. Therefore we investigated the impact of cerebral malaria on the development of reflexive pro-saccades and antisaccades. In a longitudinal study, 47 children previously admitted with retinopathy-confirmed cerebral malaria (mean age at admission 54 months), were compared with 37 local healthy controls (mean ages at first study visit 117 and 110 months respectively). In each of three or four test sessions, over a period of up to 32 months, participants completed 100 prosaccade tasks and 100 antisaccade tasks. Eye movements were recorded using infrared reflectance oculography; prosaccade, correct antisaccade and error prosaccade latency, and antisaccade directional error rate were calculated. Hierarchical linear modelling was used to investigate the effect of age and the influence of cerebral malaria on these parameters. Data were also collected from an independent, older group (mean age 183 months) of 37 local healthy participants in a separate cross-sectional study. Longitudinal data exhibited the expected decrease in latency with age for all saccade types, and a decrease in the antisaccade directional error rate. Hierarchical linear modelling confirmed that age had a statistically significant effect on all parameters (p< = 0.001). However, there were no statistically significant differences between the cerebral malaria and control groups. Combining groups, comparison with the literature demonstrated that antisaccade directional error rate for the Malawi sample was significantly higher than expected, while latencies for all saccade types were indistinguishable from published. The high directional error rate was also confirmed in the older, healthy Malawian participants from the cross sectional study. Our observation of similar oculomotor performance in cerebral malaria and control groups at long follow-up periods suggests that cerebral malaria survivors are not at a generally increased risk of persistent cognitive deficits. Our data raise questions about the prevailing hypothesis that cerebral malaria has gross impacts on the development of processes such as attention and behavioural inhibition. More importantly, our novel finding of a clear difference in antisaccade performance between all of the Malawi participants and published data suggests that the Malawian paediatric population as a whole faces serious challenges to cognitive development beyond cerebral malaria.

Grading fluorescein angiograms in malarial retinopathy.

BACKGROUND: Malarial retinopathy is an important finding in Plasmodium falciparum cerebral malaria, since it strengthens diagnostic accuracy, predicts clinical outcome and appears to parallel cerebral disease processes. Several angiographic features of malarial retinopathy have been described, but observations in different populations can only be reliably compared if consistent methodology is used to capture and grade retinal images. Currently no grading scheme exists for fluorescein angiographic features of malarial retinopathy. METHODS: A grading scheme for fluorescein angiographic images was devised based on consensus opinion of clinicians and researchers experienced in malarial retinopathy in children and adults. Dual grading were performed with adjudication of admission fluorescein images from a large cohort of children with cerebral malaria. RESULTS: A grading scheme is described and standard images are provided to facilitate future grading studies. Inter-grader agreement was >70 % for most variables. Intravascular filling defects are difficult to grade and tended to have lower inter-grader agreement (>57 %) compared to other features. CONCLUSIONS: This grading scheme provides a consistent way to describe retinal vascular damage in paediatric cerebral malaria, and can facilitate comparisons of angiographic features of malarial retinopathy between different patient groups, and analysis against clinical outcomes. Inter-grader agreement is reasonable for the majority of angiographic signs. Dual grading with expert adjudication should be used to maximize accuracy.

Developing retinal biomarkers of neurological disease: an analytical perspective.

The inaccessibility of the brain poses a problem for neuroscience. Scientists have traditionally responded by developing biomarkers for brain physiology and disease. The retina is an attractive source of biomarkers since it shares many features with the brain. Some even describe the retina as a 'window' to the brain, implying that retinal signs are analogous to brain disease features. However, new analytical methods are needed to show whether or not retinal signs really are equivalent to brain abnormalities, since this requires greater evidence than direct associations between retina and brain. We, therefore propose a new way to think about, and test, how clearly one might see the brain through the retinal window, using cerebral malaria as a case study.

Automated detection of leakage in fluorescein angiography images with application to malarial retinopathy.

The detection and assessment of leakage in retinal fluorescein angiogram images is important for the management of a wide range of retinal diseases. We have developed a framework that can automatically detect three types of leakage (large focal, punctate focal, and vessel segment leakage) and validated it on images from patients with malarial retinopathy. This framework comprises three steps: vessel segmentation, saliency feature generation and leakage detection. We tested the effectiveness of this framework by applying it to images from 20 patients with large focal leak, 10 patients with punctate focal leak, and 5,846 vessel segments from 10 patients with vessel leakage. The sensitivity in detecting large focal, punctate focal and vessel segment leakage are 95%, 82% and 81%, respectively, when compared to manual annotation by expert human observers. Our framework has the potential to become a powerful new tool for studying malarial retinopathy, and other conditions involving retinal leakage.

Automated Detection of Vessel Abnormalities on Fluorescein Angiogram in Malarial Retinopathy.

The detection and assessment of intravascular filling defects is important, because they may represent a process central to cerebral malaria pathogenesis: neurovascular sequestration. We have developed and validated a framework that can automatically detect intravascular filling defects in fluorescein angiogram images. It first employs a state-of-the-art segmentation approach to extract the vessels from images and then divide them into individual segments by geometrical analysis. A feature vector based on the intensity and shape of saliency maps is generated to represent the level of abnormality of each vessel segment. An AdaBoost classifier with weighted cost coefficient is trained to classify the vessel segments into normal and abnormal categories. To demonstrate its effectiveness, we apply this framework to 6,358 vessel segments in images from 10 patients with malarial retinopathy. The test sensitivity, specificity, accuracy, and area under curve (AUC) are 74.7%, 73.5%, 74.1% and 74.2% respectively when compared to the reference standard of human expert manual annotations. This performance is comparable to the agreement that we find between human observers of intravascular filling defects. Our method will be a powerful new tool for studying malarial retinopathy.

A comprehensive texture segmentation framework for segmentation of capillary non-perfusion regions in fundus fluorescein angiograms.

Capillary non-perfusion (CNP) in the retina is a characteristic feature used in the management of a wide range of retinal diseases. There is no well-established computation tool for assessing the extent of CNP. We propose a novel texture segmentation framework to address this problem. This framework comprises three major steps: pre-processing, unsupervised total variation texture segmentation, and supervised segmentation. It employs a state-of-the-art multiphase total variation texture segmentation model which is enhanced by new kernel based region terms. The model can be applied to texture and intensity-based multiphase problems. A supervised segmentation step allows the framework to take expert knowledge into account, an AdaBoost classifier with weighted cost coefficient is chosen to tackle imbalanced data classification problems. To demonstrate its effectiveness, we applied this framework to 48 images from malarial retinopathy and 10 images from ischemic diabetic maculopathy. The performance of segmentation is satisfactory when compared to a reference standard of manual delineations: accuracy, sensitivity and specificity are 89.0%, 73.0%, and 90.8% respectively for the malarial retinopathy dataset and 80.8%, 70.6%, and 82.1% respectively for the diabetic maculopathy dataset. In terms of region-wise analysis, this method achieved an accuracy of 76.3% (45 out of 59 regions) for the malarial retinopathy dataset and 73.9% (17 out of 26 regions) for the diabetic maculopathy dataset. This comprehensive segmentation framework can quantify capillary non-perfusion in retinopathy from two distinct etiologies, and has the potential to be adopted for wider applications.