Carol Yim Lui Cheung

Singapore Eye Research Institute, Tumasik, Singapore

Are you Carol Yim Lui Cheung?

Claim your profile

Publications (19)53.75 Total impact

  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Purpose: To investigate the determinants of pupil diameter (PD), amplitude of pupil diameter change (PD-change) and speed of pupil constriction (SPC) using video anterior segment optical coherence tomography (AS-OCT) in a population-based sample of Chinese adults. Methods: Chinese adults free from glaucoma, aged between 40 and 80 years were consecutively recruited from the population-based Singapore Chinese Eye Study. The SPC was measured by AS-OCT videography. Univariate and multivariate analyses were performed to examine the effect of demographic and ocular biometric factors (e.g., axial length [AL], anterior chamber depth [ACD], baseline PD, iris thickness at the area of the dilator muscle [ITDMR], and iris area [IA], iris bowing [IB]) on SPC, PD and PD-change. Results: A total of 266/302 (89.5%) AS-OCT videos of eligible eyes were available for analysis. Of these subjects, 64.3% were female, and the mean age of subjects (± standard deviation [SD]) was 56 ± 8.3 years. SPC was not associated with sex. In multiple regression analyses, SPC was independently associated with baseline PD (β=0.116, p=0.006). Baseline PD was independently associated with ACD (β=0.341, p<0.001), trabecular-iris space area (TISA-500, β=-4.513, ρ<0.001), IA (β=-2.796, ρ<0.001), and ITDMR (β=6.573, ρ<0.001). PD-change was independently associated with ACD (β=0.256, ρ<0.001), IA (β=-1.507, ρ<0.001), IB (β=0.630, ρ=0.011) and ITDMR (β=3.124, ρ<0.001). Conclusions: Among normal adult Chinese eyes, SPC was associated with larger baseline PDs. Larger baseline PDs and greater PD changes form dark to light were observed in deeper ACD with smaller iris cross-sectional areas and thicker irides.
    APACRS 2013; 07/2013
  • Alzheimer's and Dementia 07/2013; 9(4):P743. · 17.47 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: To examine the effect of blood pressure (BP) on retinal vascular fractal dimension (Df), a measure of microvascular network complexity and density in a multiethnic cohort. A population-based study of 3876 Chinese, Malay and Indian participants in Singapore. Retinal Df was measured using a computer-based program from digital retinal photographs. Associations between retinal Df and mean arterial BP (MABP) in the whole cohort and in each racial group were analysed using linear regression analysis. Logistic regression was used to examine the association between retinal Df and hypertension status. The mean retinal Df of the study population was 1.45 (standard deviation 0.03). After adjustment for age, sex, race, diabetes, BMI, cholesterol and creatinine levels, persons with smaller Df had higher MABP (mean difference MABP was 6.18 mmHg comparing lowest to highest Df quartiles, P < 0.001). This was similar in Chinese, Malay and Indian persons [mean difference 6.40 (P < 0.001), 4.72 (P = 0.011) and 6.62 (P < 0.001) mmHg, respectively]. Persons with smaller retinal Df were more likely to have uncontrolled treated or untreated hypertension [odds ratio 1.79 (P = 0.003) and 2.60 (P = 0.003), respectively, comparing lowest to highest Df quartiles] than those with no hypertension; this relationship was not seen comparing persons with controlled treated hypertension with no hypertension (odds ratio 1.01, P = 0.972). Hypertension was associated with a sparser retinal vascular network, which was similar across different racial/ethnic groups and most apparent in those with uncontrolled or untreated hypertension. These data suggest that microvascular remodelling can be quantified by measuring retinal vasculature.
    Journal of Hypertension 06/2013; · 4.22 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Background: Retinal vascular abnormalities are associated with cardiovascular risk factors and adverse cardiac events. Endothelial dysfunction, arterial stiffness and carotid intima-media thickening (CIMT) reflect sub-clinical atherosclerosis. In patients with autoimmune diseases, the relationship between retinal vascular changes and these vascular indices is unknown. The aim is to correlate retinal vascular changes with endothelial function, arterial stiffness and CIMT in SLE women without coronary heart disease (CHD). Methods: A single-Centre, cross-sectional study of 28 SLE women (7 with secondary anti-phospholipid syndrome; mean age 42 AE 10 years; mean disease duration 13 AE 5 years) without CHD. Retinal digital photography, brachial artery flow-mediated vasodilatation (FMD), local carotid pulse wave velocity (PWV) by a Doppler echo-tracking system and CIMT were performed. FMD was expressed as percentage change relative to baseline diameter. Results: Pearson's correlation analysis showed FMD and retinal arteriolar caliber size (RetA) was inversely correlated (r ¼ —0.47, P-value ¼ 0.02). There was no significant correlation between PWV and RetA (r ¼ —0.39, P-value ¼ 0.05) as well as between CIMT and RetA (r ¼ —0.13, P-value ¼ 0.54). Reduced FMD was associated with a wider RetA after adjusting for age and disease duration with multivariate linear regression analysis (P-¼ 0.009). FMD, PWV and CIMT were not associated with retinal venular caliber size. Conclusions: In SLE women without CHD, a wider retinal arteriolar caliber is associated with reduced FMD, independent of age or disease duration. These findings suggest retinal arteriolar caliber may be a marker of systemic endothelial dysfunction. Disclosures: The authors have declared no conflicts of interest.
    British Society of Rheumatology; 04/2013
  • [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, we present an unsupervised framework using domain priors extracted from the primary structures of the optic nerve head for automated optic cup localization. Our approach provides 3 major contributions. First, we identify a new domain prior, optic cup origin. This prior is derived from the physiological understanding that the central retinal vessels traces its origin from the optic cup before extending to the rest of the retinal. Second, we propose extracting the features of the optic nerve head from superpixels, which are obtained from low-level grouping and have more natural and descriptive features than pixel based techniques. Third, the domain knowledge comprising of optic cup origin and cup pallor, and the extracted features from superpixels are then used to drive a similarity-based label propagation and refinement scheme for the optic cup localization. Our approach was validated on a clinical online dataset, ORIGA-light, of 650 population-based images. Overall, our approach is able to achieve a 32.2% nonoverlap ratio (m1), a 33.8% relative absolute area difference (m2) and a 10.6% absolute CDR error (δ).
    Biomedical Imaging (ISBI), 2013 IEEE 10th International Symposium on; 01/2013
  • [Show abstract] [Hide abstract]
    ABSTRACT: Optic disc segmentation in retinal fundus image is important in ocular image analysis and computer aided diagnosis. Because of the presence of peripapillary atrophy which affects the deformation, it is important to have a good initialization in deformable model based optic disc segmentation. In this paper, a superpixel classification based method is proposed for the initialization. It uses histogram of superpixels from the contrast enhanced image as features. In the training, bootstrapping is adopted to handle the unbalanced cluster issue due to the presence of peripapillary atrophy. A self-assessment reliability score is computed to evaluate the quality of the initialization and the segmentation. The proposed method has been tested in a database of 650 images with optic disc boundaries marked by trained professionals manually. The experimental results show an mean overlapping error of 10.0% and standard deviation of 7.5% in the best scenario. The results also show an increase in overlapping error as the reliability score reduces, which justifies the effectiveness of the self-assessment. The method can be used for optic disc boundary initialization and segmentation in computer aided diagnosis system and the self-assessment can be used as an indicator of cases with large errors and thus enhance the usage of the automatic segmentation.
    Conference proceedings: ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Conference 08/2012; 2012:1450-3.
  • [Show abstract] [Hide abstract]
    ABSTRACT: Optic cup detection remains a challenging task in retinal image analysis, and is of particular importance for glaucoma evaluation, where disease severity is often assessed by the size of the optic cup. In this paper, we propose spatial heuristic ensembling (SHE), an approach which aims to fuse the advantages of each method based on the specific performance in each defined sector. In this way, we generate an ensembled optic cup which is obtained from the optimal combination of the component methods. We conduct experiments on the ORIGA data set of 650 retinal images and show that the ensemble approach performs better than the individual segmentations, reducing the relative overlap error, and CDR errors by as much as 0.04 CDR units. The results are promising for the continued development of such an approach for improving optic cup segmentation.
    Conference proceedings: ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Conference 08/2012; 2012:1426-9.
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, we present a new method to detect pterygiums using cornea images. Due to the similarity of appearances and spatial locations between pterygiums and cortical cataracts, pterygiums are often falsely detected as cortical cataracts on retroillumination images by a computer-aided grading system. The proposed method can be used to filter out the pterygium which improves the accuracy of cortical cataract grading system. This work has three major contributions. First, we propose a new pupil segmentation method for visible wavelength images. Second, an automatic detection method of pterygiums is proposed. Third, we develop an enhanced compute-aided cortical cataract grading system that excludes pterygiums. The proposed method is tested using clinical data and the experimental results demonstrate that the proposed method can improve the existing automatic cortical cataract grading system.
    Conference proceedings: ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Conference 08/2012; 2012:4434-7.
  • [Show abstract] [Hide abstract]
    ABSTRACT: Identifying the boundary of the optic cup excavation is of critical importance in the assessment of glaucomatous risk. Currently, most approaches are focused on the use of pallor. We present an automatic method to determine the cup excavation boundary based on vessel kinking in non-stereo retinal fundus images. The method tracks vessels using a self-initialized evolving model which adapts during propagation. A piecewise analysis of the vessel trajectory is conducted to determine vessel kinking locations, from which the optic cup is obtained. Our proposed method is validated on a large database of 618 images. The results suggest that our proposed method is able to obtain a more accurate cup excavation detection than other tested methods.
    Pattern Recognition (ICPR), 2012 21st International Conference on; 01/2012
  • [Show abstract] [Hide abstract]
    ABSTRACT: The vertical Cup-to-Disc Ratio (CDR) is an important indicator in the diagnosis of glaucoma. Automatic segmentation of the optic disc (OD) and optic cup is crucial towards a good computer-aided diagnosis (CAD) system. This paper presents a statistical model-based method for the segmentation of the optic disc and optic cup from digital color fundus images. The method combines knowledge-based Circular Hough Transform and a novel optimal channel selection for segmentation of the OD. Moreover, we extended the method to optic cup segmentation, which is a more challenging task. The system was tested on a dataset of 325 images. The average Dice coefficient for the disc and cup segmentation is 0.92 and 0.81 respectively, which improves significantly over existing methods. The proposed method has a mean absolute CDR error of 0.10, which outperforms existing methods. The results are promising and thus demonstrate a good potential for this method to be used in a mass screening CAD system.
    Computer-Based Medical Systems (CBMS), 2012 25th International Symposium on; 01/2012
  • [Show abstract] [Hide abstract]
    ABSTRACT: To investigate the effect of tumor necrosis factor alpha (TNF-α) on the mouse retinal vasculature, function, and expression of vascular endothelial growth factor-A (VEGF-A) in the retina and retinal pigment epithelium (RPE) and to evaluate the protective effect of statin therapy (fluvastatin) on retinal vascular and functional changes. A single intravenous injection of murine TNF-α (8 μg/kg body weight) was administered to one group of mice (TNF group). In the second group of mice (TNF+Statin group), a single dose of TNF-α was followed by 28 days oral medication of fluvastatin (10 mg/kg/d), and an equivalent volume of saline was administered to the third group (Control group). After 28 days, electroretinography (ERG) and fundus photography were performed. Eyes were collected for cell and molecular studies. Transcript levels of VEGF-A in retina and RPE were quantified using real-time polymerase chain reaction, and protein expression was analyzed by Western blot and immunostaining. TNF-α-injected mice showed retinal vessel tortuosity, structural change, and altered retinal function. Fluvastatin-treated mice exhibited retinal vascular, structural, and functional changes almost similar to those of the control group. VEGF-A expression was significantly upregulated in the retina and RPE of TNF-α-injected mice, and this was significantly downregulated in fluvastatin-treated mice. This study shows that the TNF-α-induced inflammatory process results in the alteration of retinal microvasculature and function, and fluvastatin could be a potential therapy for treating/preventing retinal microvascular or inflammatory complications.
    Investigative ophthalmology & visual science 08/2011; 52(10):7423-31. · 3.43 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: To compare the performance of standard automated perimetry (SAP), frequency-doubling technology (FDT) perimetry, and short-wavelength automated perimetry (SWAP) in detecting glaucoma. One hundred thirty-two eyes of 95 glaucoma patients and 37 normal subjects had retinal nerve fiber layer (RNFL) imaging and visual field testing by SAP, Matrix FDT perimetry, and Swedish interactive thresholding algorithm (SITA) SWAP at the same visit (all perimeters by Carl Zeiss Meditec, Inc., Dublin, CA). Visual field defects were confirmed with two or more consecutive examinations by the same types of perimetry. Glaucoma was defined with the reference to the RNFL thickness deviation map score (≥ 4, glaucomatous; ≤ 2, normal). The sensitivity, specificity, and area under the receiver operating characteristic curve (AUC) of MD (mean deviation) and PSD (pattern standard deviation) of the perimetries were compared. Taking all glaucoma patients into consideration, the sensitivity was highest for Matrix FDT perimetry (69%), followed by SAP (68%), and then SITA SWAP (59%). When the analysis included only patients with early glaucoma, the sensitivity decreased to 52%, 46%, and 34%, respectively, with a significant difference detected between Matrix FDT perimetry and SITA SWAP (P = 0.034). The specificity was ≥ 97% for all perimetries. The AUCs of MD and PSD followed a similar order, with Matrix FDT perimetry having the greatest AUC (0.89-0.94), followed by SAP (0.87-0.94), and then SITA SWAP (0.69-0.90). There were significant differences in sensitivities at 90% specificity between Matrix FDT perimetry and SITA SWAP (P ≤ 0.005 for MD; P ≤ 0.039 for PSD). The performance for glaucoma detection was comparable between FDT perimetry and SAP. FDT perimetry had a higher sensitivity for detecting glaucoma than did SWAP at a comparable level of specificity.
    Investigative ophthalmology & visual science 08/2011; 52(10):7325-31. · 3.43 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Glaucoma is an optic nerve disease resulting in the loss of vision. There are two common types of glaucoma: open angle glaucoma and angle closure glaucoma. Glaucoma type classification is important in glaucoma diagnosis. Clinically, ophthalmologists examine the iridocorneal angle between iris and cornea to determine the glaucoma type as well as the degree of closure. However, manual grading of the iridocorneal angle images is subjective and often time consuming. In this paper, we propose focal edge for automated iridocorneal angle grading. The iris surface is located to determine focal region and focal edges. The association between focal edges and angle grades is built through machine learning. A modified grading system with three grades is adopted. The experimental results show that the proposed method can correctly classify 87.3% open angle and 88.4% closed angle. Moreover, it can correctly classify 75.0% grade 1 and 77.4% grade 0 for angle closure cases.
    Conference proceedings: ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Conference 08/2011; 2011:4481-4.
  • [Show abstract] [Hide abstract]
    ABSTRACT: Optic disc segmentation from retinal fundus image is a fundamental but important step for automatic glaucoma diagnosis. In this paper, an optic disc segmentation method is proposed based on peripapillary atrophy elimination. The elimination is done through edge filtering, constraint elliptical Hough transform and peripapillary atrophy detection. With the elimination, edges that are likely from non-disc structures especially peripapillary atrophy are excluded to make the segmentation more accurate. The proposed method has been tested in a database of 650 images with disc boundaries marked by trained professionals manually. The experimental results by the proposed method show average m(1), m(2) and m(VD) of 10.0%, 7.4% and 4.9% respectively. It can be used to compute cup to disc ratio as well as other features for application in automatic glaucoma diagnosis systems.
    Conference proceedings: ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Conference 08/2011; 2011:6224-7.
  • [Show abstract] [Hide abstract]
    ABSTRACT: The optic nerve head (optic disc) plays an important role in the diagnosis of retinal diseases. Automatic localization and segmentation of the optic disc is critical towards a good computer-aided diagnosis (CAD) system. In this paper, we propose a method that combines edge detection, the Circular Hough Transform and a statistical deformable model to detect the optic disc from retinal fundus images. The algorithm was evaluated against a data set of 325 digital color fundus images, which includes both normal images and images with various pathologies. The result shows that the average error in area overlap is 11.3% and the average absolute area error is 10.8%, which outperforms existing methods. The result indicates a high correlation with ground truth segmentation and thus demonstrates a good potential for this system to be integrated with other retinal CAD systems.
    Conference proceedings: ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Conference 08/2011; 2011:2626-9.
  • [Show abstract] [Hide abstract]
    ABSTRACT: Min-Redundancy Max-Relevance (mRMR) is a feature selection methodology based on information theory. We explore the mRMR principle for automatic glaucoma diagnosis. Optimal candidate feature sets are acquired from a composition of clinical screening data and retinal fundus image data. An mRMR optimized classifier is further trained using the candidate feature sets to find the optimized classifier. We tested the proposed methodology on eye records of 650 subjects collected from Singapore Eye Research Institute. The experimental results demonstrate that the new classifier is much compact by using less than ¼ of the initial feature set. The ranked feature set also enables the clinicians to better access the diagnostic process of the algorithm. The work is a further step towards the advancement of the automatic glaucoma diagnosis.
    Conference proceedings: ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Conference 08/2011; 2011:6228-31.
  • Alzheimer's and Dementia 07/2011; 7(4). · 17.47 Impact Factor
  • Alzheimer's and Dementia 07/2011; 7(4). · 17.47 Impact Factor
  • Alzheimer's and Dementia 07/2011; 7(4). · 17.47 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: To evaluate the performance of progression detection and the rate of change of retinal nerve fiber layer (RNFL), neuroretinal rim, and visual field measurements in glaucoma. Prospective study. One hundred eight eyes of 70 glaucoma patients. Patients were followed up every 4 months for at least 2.9 years (median, 3.2 years) for measurement of RNFL thickness with the Stratus optical coherence tomograph (OCT) (Carl Zeiss Meditec, Dublin, CA), neuroretinal rim area with the Heidelberg Retinal Tomograph (HRT 3; Heidelberg Engineering, GmbH, Dossenheim, Germany), and visual field with the Humphrey Field Analyzer II (Carl Zeiss Meditec). Linear regression analyses were performed between visual field index (VFI), RNFL, and neuroretinal rim measurements and age, with progression defined when a significant negative trend was detected. The agreement among structural and functional measurements was evaluated with κ statistics. The mean rate of change was estimated with linear mixed modeling. The agreement on progression detection and the rate of change of RNFL, neuroretinal rim, and VFI measurements. A total of 1105 OCT, 1062 HRT, and 1099 visual field measurements were analyzed. The agreement of progression detection among the 3 investigations was poor (κ≤0.09). Ten eyes (9.3%; 9 patients) showed progression by average RNFL thickness, 16 (14.8%; 14 patients) by global neuroretinal rim area, and 35 (32.4%; 31 patients) by VFI. Only 1 eye (0.9%) had progression detected by all 3 methods. There were large variations in the rate of change of VFI, average RNFL thickness, and global neuroretinal rim area, with a range between -0.63% and -4.97% per year, -2.32% and -10.12% per year, and -0.61% and -8.48% per year, respectively. The respective mean rate estimates were -1.15% per year (95% confidence interval [CI], -1.56% to -0.73%), -0.70% per year (95% CI, -1.19% to -0.21%), and -1.06% per year (95% CI, -1.56% to -0.55%). The agreement of progression detection among RNFL, neuroretinal rim, and visual field measurements was poor, and the rate of RNFL, neuroretinal rim, and visual field progression varied considerably within and between subjects. Given this variability, interpretation of RNFL, neuroretinal rim, and VFI progression always should be evaluated on an individual basis. Proprietary or commercial disclosure may be found after the references.
    Ophthalmology 04/2011; 118(8):1551-7. · 5.56 Impact Factor