ABSTRACT
Background: As an extension of the central nervous system (CNS), the retina shares many similarities with the brain and can manifest signs of various neurological diseases, including Alzheimer's disease (AD). Objective: To investigate the retinal spectral features and develop a classification model to differentiate individuals with different brain amyloid levels. Methods: Sixty-six participants with varying brain amyloid-ß protein levels were non-invasively imaged using a hyperspectral retinal camera in the wavelength range of 450-900ânm in 5ânm steps. Multiple retina features from the central and superior views were selected and analyzed to identify their variability among individuals with different brain amyloid loads. Results: The retinal reflectance spectra in the 450-585ânm wavelengths exhibited a significant difference in individuals with increasing brain amyloid. The retinal features in the superior view showed higher inter-subject variability. A classification model was trained to differentiate individuals with varying amyloid levels using the spectra of extracted retinal features. The performance of the spectral classification model was dependent upon retinal features and showed 0.758-0.879 accuracy, 0.718-0.909 sensitivity, 0.764-0.912 specificity, and 0.745-0.891 area under curve for the right eye. Conclusions: This study highlights the spectral variation of retinal features associated with brain amyloid loads. It also demonstrates the feasibility of the retinal hyperspectral imaging technique as a potential method to identify individuals in the preclinical phase of AD as an inexpensive alternative to brain imaging.
Subject(s)
Alzheimer Disease , Amyloid beta-Peptides , Brain , Retina , Humans , Female , Male , Retina/diagnostic imaging , Alzheimer Disease/diagnostic imaging , Alzheimer Disease/metabolism , Brain/diagnostic imaging , Brain/metabolism , Aged , Amyloid beta-Peptides/metabolism , Hyperspectral Imaging , Biomarkers , Middle Aged , Aged, 80 and overABSTRACT
BACKGROUND/OBJECTIVES: Artificial intelligence can assist with ocular image analysis for screening and diagnosis, but it is not yet capable of autonomous full-spectrum screening. Hypothetically, false-positive results may have unrealized screening potential arising from signals persisting despite training and/or ambiguous signals such as from biomarker overlap or high comorbidity. The study aimed to explore the potential to detect clinically useful incidental ocular biomarkers by screening fundus photographs of hypertensive adults using diabetic deep learning algorithms. SUBJECTS/METHODS: Patients referred for treatment-resistant hypertension were imaged at a hospital unit in Perth, Australia, between 2016 and 2022. The same 45° colour fundus photograph selected for each of the 433 participants imaged was processed by three deep learning algorithms. Two expert retinal specialists graded all false-positive results for diabetic retinopathy in non-diabetic participants. RESULTS: Of the 29 non-diabetic participants misclassified as positive for diabetic retinopathy, 28 (97%) had clinically useful retinal biomarkers. The models designed to screen for fewer diseases captured more incidental disease. All three algorithms showed a positive correlation between severity of hypertensive retinopathy and misclassified diabetic retinopathy. CONCLUSIONS: The results suggest that diabetic deep learning models may be responsive to hypertensive and other clinically useful retinal biomarkers within an at-risk, hypertensive cohort. Observing that models trained for fewer diseases captured more incidental pathology increases confidence in signalling hypotheses aligned with using self-supervised learning to develop autonomous comprehensive screening. Meanwhile, non-referable and false-positive outputs of other deep learning screening models could be explored for immediate clinical use in other populations.
ABSTRACT
This paper presents a low computationally intensive and memory efficient convolutional neural network (CNN)-based fully automated system for detection of glaucoma, a leading cause of irreversible blindness worldwide. Using color fundus photographs, the system detects glaucoma in two steps. In the first step, the optic disc region is determined relying upon You Only Look Once (YOLO) CNN architecture. In the second step classification of 'glaucomatous' and 'non-glaucomatous' is performed using MobileNet architecture. A simplified version of the original YOLO net, specific to the context, is also proposed. Extensive experiments are conducted using seven state-of-the-art CNNs with varying computational intensity, namely, MobileNetV2, MobileNetV3, Custom ResNet, InceptionV3, ResNet50, 18-Layer CNN and InceptionResNetV2. A total of 6671 fundus images collected from seven publicly available glaucoma datasets are used for the experiment. The system achieves an accuracy and F1 score of 97.4% and 97.3%, with sensitivity, specificity, and AUC of respectively 97.5%, 97.2%, 99.3%. These findings are comparable with the best reported methods in the literature. With comparable or better performance, the proposed system produces significantly faster decisions and drastically minimizes the resource requirement. For example, the proposed system requires 12 times less memory in comparison to ResNes50, and produces 2 times faster decisions. With significantly less memory efficient and faster processing, the proposed system has the capability to be directly embedded into resource limited devices such as portable fundus cameras.
Subject(s)
Glaucoma , Optic Disk , Humans , Glaucoma/diagnostic imaging , Optic Disk/diagnostic imaging , Fundus Oculi , Neural Networks, Computer , Diagnostic Techniques, OphthalmologicalABSTRACT
INTRODUCTION: Automated assessment of age-related macular degeneration (AMD) using optical coherence tomography (OCT) has gained significant research attention in recent years. Though a list of convolutional neural network (CNN)-based methods has been proposed recently, methods that uncover the decision-making process of CNNs or critically interpret CNNs' decisions in the context are scant. This study aims to bridge this research gap. METHODS: We independently trained several state-of-the-art CNN models, namely, VGG16, VGG19, Xception, ResNet50, InceptionResNetV2 for AMD detection and applied CNN visualization techniques, namely, Grad-CAM, Grad-CAM++, Score CAM, Faster Score CAM to highlight the regions of interest utilized by the CNNs in the context. Retinal layer segmentation methods were also developed to explore how the CNN regions of interest related to the layers of the retinal structure. Extensive experiments involving 2130 SD-OCT scans collected from Duke University were performed. RESULTS: Experimental analysis shows that Outer Nuclear Layer to Inner Segment Myeloid (ONL-ISM) influences the AMD detection decision heavily as evident from the normalized intersection (NI) scores. For AMD cases the obtained average NI scores were respectively 13.13%, 17.2%, 9.7%, 10.95%, and 11.31% for VGG16, VGG19, ResNet50, Xception, and Inception ResNet V2, whereas, for normal cases, these values were respectively 21.7%, 21.3%, 16.85%, 10.175% and 16%. CONCLUSION: Critical analysis reveals that the ONL-ISM is the most contributing layer in determining AMD, followed by Nerve Fiber Layer to Inner Plexiform Layer (NFL-IPL).
ABSTRACT
The aim of the study was to assess various retinal vessel parameters of diabetes mellitus (DM) patients and their correlations with systemic factors in type 2 DM. A retrospective exploratory study in which 21 pairs of baseline and follow-up images of patients affected by DM were randomly chosen from the Sankara Nethralaya−Diabetic Retinopathy Study (SN DREAMS) I and II datasets. Patients' fundus was photographed, and the diagnosis was made based on Klein classification. Vessel thickness parameters were generated using a web-based retinal vascular analysis platform called VASP. The thickness changes between the baseline and follow-up images were computed and normalized with the actual thicknesses of baseline images. The majority of parameters showed 10~20% changes over time. Vessel width in zone C for the second vein was significantly reduced from baseline to follow-up, which showed positive correlations with systolic blood pressure and serum high-density lipoproteins. Fractal dimension for all vessels in zones B and C and fractal dimension for vein in zones A, B and C showed a minimal increase from baseline to follow-up, which had a linear relationship with diastolic pressure, mean arterial pressure, serum triglycerides (p < 0.05). Lacunarity for all vessels and veins in zones A, B and C showed a minimal decrease from baseline to follow-up which had a negative correlation with pulse pressure and positive correlation with serum triglycerides (p < 0.05). The vessel widths for the first and second arteries significantly increased from baseline to follow-up and had an association with high-density lipoproteins, glycated haemoglobin A1C, serum low-density lipoproteins and total serum cholesterol. The central reflex intensity ratio for the second artery was significantly decreased from baseline to follow-up, and positive correlations were noted with serum triglyceride, serum low-density lipoproteins and total serum cholesterol. The coefficients for branches in zones B and C artery and the junctional exponent deviation for the artery in zone A decreased from baseline to follow-up showed positive correlations with serum triglycerides, serum low-density lipoproteins and total serum cholesterol. Identifying early microvascular changes in diabetic patients will allow for earlier intervention, improve visual outcomes and prevent vision loss.
ABSTRACT
OBJECTIVE: We analyzed whether any change in capillary density in the retinal circulation could be detected in patients with hypertension in the prediabetic stage. RESEARCH DESIGN AND METHODS: In a cross-sectional analysis, we assessed capillary density in the foveal (CDF) and parafoveal retinal areas using optical coherence tomography-angiography in 62 patients with hypertension and normal glucose metabolism and 40 patients with hypertension and prediabetes. RESULTS: The CDF was lower in patients with prediabetes than in those with normal glucose metabolism. Moreover, we found a correlation between CDF and HbA1c and glucose levels for the entire cohort. In patients with HbA1c <6.5% (48 mmol/mol), CDF was lower in patients with HOMA for insulin resistance (HOMA-IR) ≥2.5 than in patients with HOMA-IR <2.5. CONCLUSIONS: Patients with hypertension and prediabetes display retinal capillary changes, and an association with markers of glucose metabolism exists, even within a nondiabetic HbA1c range.
Subject(s)
Hypertension , Insulin Resistance , Prediabetic State , Blood Glucose/metabolism , Cross-Sectional Studies , Glucose , Glycated Hemoglobin/metabolism , Humans , Hypertension/complications , Prediabetic State/complicationsABSTRACT
BACKGROUND: A nocturnal non-dipping pattern has been associated with hypertension-mediated organ damage (HMOD), morbidity and mortality. Retinal imaging through application of modern technologies including optical coherence tomography angiography (OCT-A) can provide detailed insights into early vascular damage. In this observational study, we investigated the relationship of microscopic vascular density in the retina measured with OCT-A and nocturnal blood pressure (BP) dipping. METHODS: Retinal OCT-A and ambulatory BP monitoring (ABPM) data prospectively obtained from 142 patients referred to a tertiary hypertension clinic were analysed with regression models for associations between BP night-time dipping and retinal capillary vascular density in three different zones around the fovea. RESULTS: More pronounced nocturnal SBP and DBP dipping was significantly associated with increased vascular density in the central foveal area of the retina. These associations were robust to adjustment for other available risk factors including mean daytime BP. Parafoveal and whole image vascular density did not show equivalent significant associations with nocturnal BP dipping. The results were reproducible when assessed in a subgroup of patients who had concomitant type 2 diabetes. CONCLUSION: Foveal vascular density was associated with the nocturnal BP dipping pattern in hypertensive patients. These associations were robust to adjustment of relevant factors such as daytime BP. Our findings highlight the importance of nocturnal BP features reflected in ambulatory BP monitoring in the assessment of HMOD. Whether routine assessment of retinal damage markers may improve risk management of hypertensive patients remains to be determined.
Subject(s)
Diabetes Mellitus, Type 2 , Hypertension , Blood Pressure , Blood Pressure Monitoring, Ambulatory , Circadian Rhythm , Humans , Hypertension/complications , Microvascular Density , Retina/diagnostic imagingABSTRACT
Microvascular disease and rarefaction are key pathological hallmarks of hypertension. The retina uniquely allows direct, non-invasive investigation of the microvasculature. Recently developed optical coherence tomography angiography now allows investigation of the fine retinal capillaries, which may provide a superior marker of overall vascular damage. This was a prospective cross-sectional study to collect retinal capillary density data on 300 normal eyes from 150 hypertensive adults, and to investigate possible associations with other organ damage markers. The average age of participants was 54 years and there was a greater proportion of males (85; 57%) than females. Multivariate, confounder adjusted linear regression showed that retinal capillary rarefaction in the parafovea was associated with increased pulse wave velocity (ß = - 0.4, P = 0.04), log-albumin/creatinine ratio (ß = - 0.71, P = 0.003), and with reduced estimated glomerular filtration rate (ß = 0.04, P = 0.02). Comparable significant associations were also found for whole-image vascular-density, for foveal vascular-density significant associations were found with pulse wave velocity and estimated glomerular filtration rate only. Our results indicate that retinal capillary rarefaction is associated with arterial stiffness and impaired kidney function. Retinal capillary rarefaction may represent a useful and simple test to assess the integrated burden of hypertension on the microvasculature irrespective of current blood pressure levels.
Subject(s)
Arteries/pathology , Capillaries/pathology , Fovea Centralis/pathology , Hypertension/pathology , Kidney/pathology , Microvascular Rarefaction/pathology , Retinal Vessels/pathology , Blood Pressure/physiology , Cross-Sectional Studies , Female , Fluorescein Angiography/methods , Glomerular Filtration Rate/physiology , Humans , Male , Microcirculation/physiology , Microvessels/pathology , Middle Aged , Prospective Studies , Pulse Wave Analysis/methods , Tomography, Optical Coherence/methods , Vascular Stiffness/physiologyABSTRACT
Patients with progressing chronic kidney disease (CKD) are more likely to experience cardio- and cerebrovascular events than progressing to end-stage renal disease. The authors explored whether retinal microvascular calibers differed with the degree of renal impairment and between the standard and extended optic disk and may serve as a simple additional tool for risk stratification in this highly vulnerable patient cohort. The authors analyzed central retinal arteriolar and venular equivalent calibers (CRAE, CRVE) at different retinal zones (zone B&C) using digital retinal imaging in hypertensive patients with stage 2 (n = 66) or stage 3 CKD (n = 30). Results were adjusted for age, sex, HbA1c, and 24-hour diastolic blood pressure. Mean eGFR was 77.7 ± 8.9 and 48.8 ± 7.9 ml/min/1.73 m2 for stage 2 and 3 CKD, respectively. CRAE and CRVE in zones B and C were significantly lower in patients with stage 3 CKD compared to patients with stage 2 CKD (CRAE-B:141.1 ± 21.4 vs. 130.5 ± 18.9 µm, p = .030; CRAE-C:137.4 ± 19.4 vs 129.2 ± 18.2 µm, p = .049; CRVE-B:220.8 ± 33.0 vs. 206.0 ± 28.4 µm, p = .004; and CRVE-C:215.9 ± 33.0 vs. 201.2 ± 25.1µm, p = .003). In patients with stage 2 CKD, CRAE-B was higher than CRAE-C (141.1 ± 21.4 vs. 137.4 ± 19.4µm, p < .001). In contrast, such a difference was not found in patients with stage 3 CKD. CRAE of both retinal zones correlated with eGFR for the entire cohort. In patients with stage 3 CKD, retinal narrowing is more pronounced compared to patients with stage 2 CKD. Whether the novel observation of difference in arteriolar caliber between zones B and C in stage 2 CKD could serve as an early marker of CKD progression warrants further investigation.
Subject(s)
Hypertension , Renal Insufficiency, Chronic , Arterioles , Glomerular Filtration Rate , Humans , Renal Insufficiency, Chronic/diagnosis , Renal Insufficiency, Chronic/epidemiology , Retinal Vessels/diagnostic imagingABSTRACT
Night-time blood pressure (BP) is an important predictor of cardiovascular outcomes. Its assessment, however, remains challenging due to limited accessibility to ambulatory BP devices in many settings, costs, and other factors. We hypothesized that BP measured in a supine position during daytime may perform similarly to night-time BP when modeling their association with vascular hypertension-mediated organ damage (HMOD). Data from 165 hypertensive patients were used who as part of their routine clinic workup had a series of standardized BP measurements including seated attended office, seated and supine unattended office, and ambulatory BP monitoring. HMOD was determined by assessment of kidney function and pulse wave velocity. Correlation analysis was carried out, and univariate and multivariate models were fitted to assess the extent of shared variance between the BP modalities and their individual and shared contribution to HMOD variables. Of all standard non-24-hour systolic BP assessments, supine systolic BP shared the highest degree of variance with systolic night-time BP. In univariate analysis, both systolic supine and night-time BP were strong determinants of HMOD variables. In multivariate models, supine BP outperformed night-time BP as the most significant determinant of HMOD. These findings indicate that supine BP may not only be a clinically useful surrogate for night-time BP when ambulatory BP monitoring is not available, but also highlights the possibility that unattended supine BP may be more closely related to HMOD than other BP measurement modalities, a proposition that requires further investigations in prospective studies.
Subject(s)
Hypertension , Pulse Wave Analysis , Blood Pressure , Blood Pressure Monitoring, Ambulatory , Humans , Hypertension/diagnosis , Prospective StudiesABSTRACT
PURPOSE: There has been increasing interest in identifying non-invasive, imaging biomarkers for neurodegenerative disorders of the central nervous system (CNS). The aim of this proof-of-concept study was to investigate whether corneal sensory nerve and dendritic cell (DC) parameters, captured using in vivo confocal microscopy (IVCM), are altered in individuals with mild cognitive impairment (MCI) and Alzheimer's disease (AD). METHODS: Fifteen participants were recruited from the Australian Imaging Biomarkers and Lifestyle (AIBL) study in Melbourne, VIC, Australia. The cohort consisted of cognitively normal (CN) individuals (n = 5), and those with MCI (n = 5) and AD (n = 5). Participants underwent a slit lamp examination of the anterior segment, followed by corneal imaging using laser-scanning in vivo confocal microscopy (IVCM) of the central and inferior whorl regions. Corneal DC density, field area, perimeter, circularity index, aspect ratio, and roundness were quantified using Image J. Quantitative data were derived for corneal nerve parameters, including nerve fiber length (CNFL), fiber density (CNFD), branch density (CNBD), and diameter. RESULTS: Corneal DC field area and perimeter were greater in individuals with MCI, relative to CN controls, in both the central and inferior whorl regions (p < 0.05 for all comparisons). In addition, corneal DCs in the whorl region of MCI eyes had lower circularity and roundness indices and a higher aspect ratio relative to CNs (p < 0.05 for all comparisons). DC density was similar across participant groups in both corneal regions. There was a trend toward lower quantitative parameters for corneal nerve architecture in the AD and MCI groups compared with CN participants, however, the inter-group differences did not reach statistical significance. Central corneal nerve diameters were similar between groups. CONCLUSION: This study is the first to report morphological differences in corneal DCs in humans with MCI. These differences were evident in both the central and mid-peripheral cornea, and in the absence of significant nerve abnormalities or a difference in DC density. These findings justify future large-scale studies to assess the utility of corneal IVCM and DC analysis for identifying early stage pathology in neurodegenerative disorders of the CNS.
ABSTRACT
Stargardt disease is one of the most common forms of inherited retinal disease and leads to permanent vision loss. A diagnostic feature of the disease is retinal flecks, which appear hyperautofluorescent in fundus autofluorescence (FAF) imaging. The size and number of these flecks increase with disease progression. Manual segmentation of flecks allows monitoring of disease, but is time-consuming. Herein, we have developed and validated a deep learning approach for segmenting these Stargardt flecks (1750 training and 100 validation FAF patches from 37 eyes with Stargardt disease). Testing was done in 10 separate Stargardt FAF images and we observed a good overall agreement between manual and deep learning in both fleck count and fleck area. Longitudinal data were available in both eyes from 6 patients (average total follow-up time 4.2 years), with both manual and deep learning segmentation performed on all (n = 82) images. Both methods detected a similar upward trend in fleck number and area over time. In conclusion, we demonstrated the feasibility of utilizing deep learning to segment and quantify FAF lesions, laying the foundation for future studies using fleck parameters as a trial endpoint.
Subject(s)
Stargardt Disease/pathology , Zinc Phosphate Cement/metabolism , Adolescent , Adult , Aged , Aged, 80 and over , Child , Deep Learning , Electroretinography/methods , Female , Fluorescein Angiography/methods , Fundus Oculi , Humans , Longitudinal Studies , Male , Middle Aged , Ophthalmoscopy/methods , Optical Imaging/methods , Retina/metabolism , Retina/pathology , Retinal Pigment Epithelium/metabolism , Retinal Pigment Epithelium/pathology , Stargardt Disease/metabolism , Tomography, Optical Coherence/methods , Young AdultABSTRACT
Purpose: The retinal blood vessels reflect changes in the brain's micro-circulation and these changes have been shown to correlate with the incidence of diseases such as stroke, heart disease and Alzheimer's disease. Studies investigating the retinal vasculature routinely use pupil dilation with tropicamide to optimize image acquisition and quality. The aim of this study was to investigate the effects of tropicamide on retinal vascular parameters using retinal photography.Methods: The study was performed on 41 healthy young subjects of both sexes, using tropicamide to dilate only the right pupil, leaving the left as a control.Results: Pupil dilation with tropicamide resulted in reduced retinal vessel width measures based on standardized approaches, particularly reduced arteriolar caliber (p < .0005). However, closer investigation of the images revealed reduced fundus image magnification in the post-tropicamide images, based on reduced optic nerve head diameter (p < .0005) and longitudinal analysis with image registration and affine transformation (p < .0001). No change in vessel width parameters was observed after adjustment for image magnification.Conclusion: These results suggest that tropicamide does not change the width of the retinal vessels, however width parameters as measured by standard approaches may be reduced due to image magnification changes resulting from cycloplegia. In this study, improved optic nerve head segmentation for image scale conversion removed the magnification error. With this correction, the tropicamide intervention had no effect on vessel width parameters in young healthy people and could be utilized in future without affecting the results of retinal vascular analysis.
Subject(s)
Mydriatics/pharmacology , Retinal Vessels/drug effects , Tropicamide/pharmacology , Adult , Female , Humans , Male , Ophthalmic Solutions/pharmacology , Pupil/drug effects , Young AdultABSTRACT
Glaucoma is rated as the leading cause of irreversible vision loss worldwide. Early detection of glaucoma is important for providing timely treatment and minimizing the vision loss. In this paper, we developed a robust segmentation method for optic disc and cup segmentation using a modified U-Net architecture, which combines the widely adopted pre-trained ResNet-34 model as encoding layers with classical U-Net decoding layers. The model was trained on the newly available RIGA dataset, and achieved an average dice value of 97.31% for disc segmentation and 87.61% for cup segmentation, comparable to that of the experts' performance for optic disc/cup segmentation and Cup-Disc-Ratio (CDR) calculation on a reserved RIGA dataset. When tested on DRISHTI-GS and RIM-ONE dataset without re-training or fine-tuning, the model achieved comparable performance to that of the state-of-the-art in literature. We have also fine-tuned the model on two databases, which achieves an average disc dice value of 97.38% and cup dice value of 88.77% for DRISHTI-GS test set, disc dice of 96.10% and cup dice of 84.45% for RIM-ONE database, which is the state-of-the-art performance on both databases in terms of cup dice and disc dice value. The advantage of the proposed method is the combination of the pre-trained ResNet and U-Net, which avoids training the network from scratch, thereby enabling fast network training with less epochs, thus further avoids over-fitting and achieves robust performance.
Subject(s)
Deep Learning , Glaucoma/diagnosis , Optic Disk/diagnostic imaging , Databases, Factual , HumansABSTRACT
Globally, Alzheimer's disease (AD) is a growing health and economic challenge that has no effective cure. Recent clinical trials indicate that preclinical treatment may be required but a routine screening tool for AD has been elusive. Hence, a simple, yet sensitive biomarker for preclinical AD, when the disease is most likely to be amenable to treatment, is lacking. Due to several features, the eye has been explored for this purpose and, among the ocular tissues, the retina has received the most attention. Currently, major works investigating the potential AD diagnosis by detecting amyloid-ß (Aß) signatures in the retinal tissue are underway, while the anterior eye is more accessible for in vivo imaging and examination. This report provides a concise review of current literature on the anterior eye components, including the crystalline lens, cornea, and aqueous humor, in AD. We also discuss the potential for assessment of the corneal nerve structure and regeneration as well as conjunctival tissue for AD-related alterations. The crystalline lens has received considerable attention, but further research is required to confirm whether Aß accumulates in the lens and whether it mirrors brain neuropathologic changes, particularly in preclinical AD. The rich corneal neural network and conjunctival vasculature also merit exploration in future studies to shed light on their potential association with AD pathologic changes.
Subject(s)
Alzheimer Disease/diagnosis , Alzheimer Disease/metabolism , Amyloid beta-Peptides/metabolism , Anterior Eye Segment/metabolism , Biomarkers/metabolism , Animals , Aqueous Humor/metabolism , Cornea/metabolism , Humans , Lens, Crystalline/metabolismABSTRACT
Worldwide there are over 46 million people living with dementia, and this number is expected to double every 20 years reaching about 131 million by 2050. The cost to the community and government health systems, as well as the stress on families and carers is incalculable. Over three decades of research into this disease have been undertaken by several research groups in Australia, including work by our original research group in Western Australia which was involved in the discovery and sequencing of the amyloid-ß peptide (also known as Aß or A4 peptide) extracted from cerebral amyloid plaques. This review discusses the journey from the discovery of the Aß peptide in Alzheimer's disease (AD) brain to the establishment of pre-clinical AD using PET amyloid tracers, a method now serving as the gold standard for developing peripheral diagnostic approaches in the blood and the eye. The latter developments for early diagnosis have been largely achieved through the establishment of the Australian Imaging Biomarker and Lifestyle research group that has followed 1,100 Australians for 11 years. AIBL has also been instrumental in providing insight into the role of the major genetic risk factor apolipoprotein E É4, as well as better understanding the role of lifestyle factors particularly diet, physical activity and sleep to cognitive decline and the accumulation of cerebral Aß.
Subject(s)
Alzheimer Disease/diagnosis , Alzheimer Disease/prevention & control , Alzheimer Disease/epidemiology , Alzheimer Disease/metabolism , Amyloidogenic Proteins/metabolism , Animals , Australia/epidemiology , Biomarkers/metabolism , Humans , Oxidative Stress/physiologyABSTRACT
In this paper we systematically evaluate the performance of several state-of-the-art local feature detectors and descriptors in the context of longitudinal registration of retinal images. Longitudinal (temporal) registration facilitates to track the changes in the retina that has happened over time. A wide number of local feature detectors and descriptors exist and many of them have already applied for retinal image registration, however, no comparative evaluation has been made so far to analyse their respective performance. In this manuscript we evaluate the performance of the widely known and commonly used detectors such as Harris, SIFT, SURF, BRISK, and bifurcation and cross-over points. As of descriptors SIFT, SURF, ALOHA, BRIEF, BRISK and PIIFD are used. Longitudinal retinal image datasets containing a total of 244 images are used for the experiment. The evaluation reveals some potential findings including more robustness of SURF and SIFT keypoints than the commonly used bifurcation and cross-over points, when detected on the vessels. SIFT keypoints can be detected with a reliability of 59% for without pathology images and 45% for with pathology images. For SURF keypoints these values are respectively 58% and 47%. ALOHA descriptor is best suited to describe SURF keypoints, which ensures an overall matching accuracy, distinguishability of 83%, 93% and 78%, 83% for without pathology and with pathology images respectively.
Subject(s)
Image Interpretation, Computer-Assisted/methods , Pattern Recognition, Automated/methods , Retina/physiopathology , Retinoscopy/methods , Algorithms , Humans , Reproducibility of Results , Time FactorsABSTRACT
Cortical cholinergic deficiency is prominent in Alzheimer's disease (AD), and published findings of diminished pupil flash response in AD suggest that this deficiency may extend to the visual cortical areas and anterior eye. Pupillometry is a low-cost, noninvasive technique that may be useful for monitoring cholinergic deficits which generally lead to memory and cognitive disorders. The aim of the study was to evaluate pupillometry for early detection of AD by comparing the pupil flash response (PFR) in AD (N = 14) and cognitively normal healthy control (HC, N = 115) participants, with the HC group stratified according to high (N = 38) and low (N = 77) neocortical amyloid burden (NAB). Constriction phase PFR parameters were significantly reduced in AD compared to HC (maximum acceleration p < 0.05, maximum velocity p < 0.0005, average velocity p < 0.005, and constriction amplitude p < 0.00005). The high-NAB HC subgroup had reduced PFR response cross-sectionally, and also a greater decline longitudinally, compared to the low-NAB subgroup, suggesting changes to pupil response in preclinical AD. The results suggest that PFR changes may occur in the preclinical phase of AD. Hence, pupillometry has a potential as an adjunct for noninvasive, cost-effective screening for preclinical AD.
ABSTRACT
OBJECTIVE: This study investigated the retinal arteriolar central reflex (CR, the central reflection observed in photographs of retinal vessels), which may provide information about micro-vascular health in the retina and also the brain, due to the homology between these vascular networks. The study also describes a novel computer based semi-automated technique that accurately quantifies retinal arteriolar CR and vessel width, and calculates the CR to vessel width ratio (CRR) from digital retinal photographs. METHODS: Digital retinal photographs were collected from participants in the Australian Imaging, Biomarkers and Lifestyle study of ageing (AIBL), including 25 participants diagnosed with Alzheimer's disease (AD) (age 72.4 ± 7.5 yrs, 12 male, 13 female) and 123 elderly participants without dementia (cognitively normals: CN) (age 71.6 ± 5.6 yrs, 55 male, 68 female). Using a sub-cohort of 144 (22 AD, 122 CN) with the novel CRR measures, we identified significantly higher CRR levels in AD participants (mean CRR 0.253 (SD 0.04)) as compared with CN's (mean CRR 0.231 (SD 0.04), p = 0.025). Adjustment for APOE ε4 allele status however, reduced the significance (p = 0.081). CRR was significantly higher in APOE ε4 allele carriers (mean CRR 0.254 (SD 0.03) as compared with non-carriers (mean CRR 0.224 (SD 0.05), p < 0.0001). RESULTS: These data indicate that CRR is strongly linked to APOE ε4 status and exhibits a weaker, independent trend with AD diagnosis. The retina may be useful as a novel model for non-invasive monitoring of the effects of APOE ε4 on the central nervous system, particularly in cerebrovascular disease.
Subject(s)
Alzheimer Disease/genetics , Alzheimer Disease/pathology , Apolipoprotein E4/genetics , Arterioles/pathology , Retinal Vessels/pathology , Aged , Aging/genetics , Alzheimer Disease/diagnostic imaging , Arterioles/diagnostic imaging , Arterioles/physiology , Arterioles/physiopathology , Cohort Studies , Female , Genetic Association Studies , Genotype , Heterozygote , Humans , Image Processing, Computer-Assisted/methods , Male , Organ Size , Pattern Recognition, Automated , Photography , Retinal Vessels/anatomy & histology , Retinal Vessels/diagnostic imaging , Retinal Vessels/physiopathologyABSTRACT
BACKGROUND: Improvements in acute care of stroke patients have decreased mortality, but survivors are still at increased risk of future vascular events and mitigation of this risk requires thorough assessment of the underlying factors leading to the stroke. The brain and eye share a common embryological origin and numerous similarities exist between the small vessels of the retina and brain. Recent population-based studies have demonstrated a close link between retinal vascular changes and stroke, suggesting that retinal photography could have utility in assessing underlying stroke risk factors and prognosis after stroke. Modern imaging equipment can facilitate precise measurement and monitoring of vascular features. However, use of this equipment is a challenge in the stroke ward setting as patients are frequently unable to maintain the required seated position, and pupil dilatation is often not feasible as it could potentially obscure important neurological signs of stroke progression. MATERIALS AND METHODS: This small study investigated the utility of a novel handheld, nonmydriatic retinal camera in the stroke ward and explored associations between retinal vascular features and stroke risk factors. This camera circumvented the practical limitations of conducting retinal photography in the stroke ward setting. RESULTS: A positive correlation was found between carotid disease and both mean width of arterioles (r = .40, P = .00571) and venules (r = .30, P = .0381). CONCLUSIONS: The results provide further evidence that retinal vascular features are clinically informative about underlying stroke risk factors and demonstrate the utility of handheld retinal photography in the stroke ward.