Ophthalmology (Eye & Ear Hospital) - Research Publications

Permanent URI for this collection

Search Results

Now showing 1 - 10 of 14
  • Item
    Thumbnail Image
    A Deep Learning System for Fully Automated Retinal Vessel Measurement in High Throughput Image Analysis.
    Shi, D ; Lin, Z ; Wang, W ; Tan, Z ; Shang, X ; Zhang, X ; Meng, W ; Ge, Z ; He, M (Frontiers Media SA, 2022)
    MOTIVATION: Retinal microvasculature is a unique window for predicting and monitoring major cardiovascular diseases, but high throughput tools based on deep learning for in-detail retinal vessel analysis are lacking. As such, we aim to develop and validate an artificial intelligence system (Retina-based Microvascular Health Assessment System, RMHAS) for fully automated vessel segmentation and quantification of the retinal microvasculature. RESULTS: RMHAS achieved good segmentation accuracy across datasets with diverse eye conditions and image resolutions, having AUCs of 0.91, 0.88, 0.95, 0.93, 0.97, 0.95, 0.94 for artery segmentation and 0.92, 0.90, 0.96, 0.95, 0.97, 0.95, 0.96 for vein segmentation on the AV-WIDE, AVRDB, HRF, IOSTAR, LES-AV, RITE, and our internal datasets. Agreement and repeatability analysis supported the robustness of the algorithm. For vessel analysis in quantity, less than 2 s were needed to complete all required analysis.
  • Item
    Thumbnail Image
    Retinal age gap as a predictive biomarker of future risk of Parkinson's disease
    Hu, W ; Wang, W ; Wang, Y ; Chen, Y ; Shang, X ; Liao, H ; Huang, Y ; Bulloch, G ; Zhang, S ; Kiburg, K ; Zhang, X ; Tang, S ; Yu, H ; Yang, X ; He, M ; Zhu, Z (OXFORD UNIV PRESS, 2022-03-01)
    INTRODUCTION: retinal age derived from fundus images using deep learning has been verified as a novel biomarker of ageing. We aim to investigate the association between retinal age gap (retinal age-chronological age) and incident Parkinson's disease (PD). METHODS: a deep learning (DL) model trained on 19,200 fundus images of 11,052 chronic disease-free participants was used to predict retinal age. Retinal age gap was generated by the trained DL model for the remaining 35,834 participants free of PD at the baseline assessment. Cox proportional hazards regression models were utilised to investigate the association between retinal age gap and incident PD. Multivariable logistic model was applied for prediction of 5-year PD risk and area under the receiver operator characteristic curves (AUC) was used to estimate the predictive value. RESULTS: a total of 35,834 participants (56.7 ± 8.04 years, 55.7% female) free of PD at baseline were included in the present analysis. After adjustment of confounding factors, 1-year increase in retinal age gap was associated with a 10% increase in risk of PD (hazard ratio [HR] = 1.10, 95% confidence interval [CI]: 1.01-1.20, P = 0.023). Compared with the lowest quartile of the retinal age gap, the risk of PD was significantly increased in the third and fourth quartiles (HR = 2.66, 95% CI: 1.13-6.22, P = 0.024; HR = 4.86, 95% CI: 1.59-14.8, P = 0.005, respectively). The predictive value of retinal age and established risk factors for 5-year PD risk were comparable (AUC = 0.708 and 0.717, P = 0.821). CONCLUSION: retinal age gap demonstrated a potential for identifying individuals at a high risk of developing future PD.
  • Item
    Thumbnail Image
    Body mass index is not associated with early onset cataract in the 45 and Up cohort study
    Zhang, J ; Wang, W ; Yang, G ; Ha, J ; Tan, X ; Shang, X ; Zhu, Z ; Han, X ; Liu, Z ; Zhang, L ; He, M ; Luo, L (AME PUBLISHING COMPANY, 2021-11)
    BACKGROUND: Body mass index (BMI) has been reported to be associated with age-related cataract, whereas its impact on early onset cataract (EOC) remains unknown. METHODS: A total of 73,007 individuals aged 45-55 years who had no previous cataract surgeries at baseline were enrolled from the population-based 45 and Up Study. BMI was calculated based on self-reported height and weight from the baseline questionnaire. Data on cataract surgeries were obtained from the Medicare Benefits Schedule database. EOC was defined as cataract surgically treated prior to 65 years of age. A Cox proportional hazards regression was used to assess the association between BMI and the incidence of EOC during the follow-up. RESULTS: A total of 1,764 participants underwent cataract surgery over 643,717 person-years of follow-up. No significant association was observed between BMI and EOC (P for trend 0.35). Among participants who drank 5 to 7 alcoholic drinks per week, a 73% and 27% reduction in the risk of EOC was observed in participants with a BMI of 18.5-19.99 and 25.0-27.49 kg/m2, respectively, compared to those with a BMI of 20.0-22.49 kg/m2. CONCLUSIONS: No association was identified between BMI and the incidence of EOC. Moderate alcohol intake may be protective against EOC.
  • Item
    Thumbnail Image
    Spatial Analysis of Incidence of Diagnosed Type 2 Diabetes Mellitus and Its Association With Obesity and Physical Inactivity
    Wu, J ; Wang, Y ; Xiao, X ; Shang, X ; He, M ; Zhang, L (FRONTIERS MEDIA SA, 2021-10-28)
    OBJECTIVES: To investigate the spatial distribution of 10-year incidence of diagnosed type 2 diabetes mellitus (T2DM) and its association with obesity and physical inactivity at a reginal level breakdown. METHODS: Demographic, behavioral, medical and pharmaceutical and diagnosed T2DM incidence data were collected from a cohort of 232,064 participants who were free of diabetes at enrolment in the 45 and Up Study, conducted in the state of New South Wales (NSW), Australia. We examined the geographical trend and correlation between obesity prevalence, physical inactivity rate and age-and-gender-adjusted cumulative incidence of T2DM, aggregated based on geographical regions. RESULT: The T2DM incidence, prevalence of obesity and physical inactivity rate at baseline were 6.32%, 20.24%, and 18.7%, respectively. The spatial variation of T2DM incidence was significant (Moran's I=0.52; p<0.01), with the lowest incidence of 2.76% in Richmond Valley-Coastal and the highest of 12.27% in Mount Druitt. T2DM incidence was significantly correlated with the prevalence of obesity (Spearman r=0.62, p<0.001), percentage of participants having five sessions of physical activities or less per week (r=0.79, p<0.001) and percentage of participants walked to work (r=-0.44, p<0.001). The geographical variations in obesity prevalence and physical inactivity rate resembled the geographical variation in the incidence of T2DM. CONCLUSION: The spatial distribution of T2DM incidence is significantly associated with the geographical prevalence of obesity and physical inactivity rate. Regional campaigns advocating the importance of physical activities in response to the alarming T2DM epidemic should be promoted.
  • Item
    Thumbnail Image
    Impact of Retinopathy and Systemic Vascular Comorbidities on All-Cause Mortality.
    Zhu, Z ; Shang, X ; Wang, W ; Ha, J ; Chen, Y ; He, J ; Yang, X ; He, M (Frontiers Media SA, 2021)
    PURPOSE: To assess the impact of retinopathy and systemic vascular comorbidities on the all-cause mortality in a representative U.S. sample. METHODS: A total of 5703 participants (≥40 years old) from the 2005-2008 National Health and Nutrition Examination Survey. The Early Treatment Diabetic Retinopathy Study grading scale was used to evaluate the retinopathy status. Systemic vascular comorbidities included diabetes mellitus (DM), high blood pressure (HBP), chronic kidney disease (CKD) and cardiovascular disease (CVD). Time to death was calculated as the time from baseline to either the date of death or censoring (December 31st, 2015), whichever came first. Risks of mortality were estimated using Cox proportional hazards models after adjusting for confounders and vascular comorbidities. RESULTS: After a median follow-up of 8.33 years (IQR: 7.50-9.67 years), there were 949 (11.8%) deaths from all causes. After adjusting for confounders, the presence of retinopathy predicted higher all-cause mortality (hazard ratio (HR), 1.41; 95% confidence interval (CI), 1.08-1.83). The all-cause mortality among participants with both retinopathy and systemic vascular comorbidities including DM (HR, 1.72; 95% CI, 1.21-2.43), HBP (HR, 1.47; 95% CI, 1.03-2.10), CKD (HR, 1.73; 95% CI, 1.26-2.39) and CVD (HR, 1.92; 95% CI, 1.21-3.04) was significantly higher than that among those without either condition. When stratified by diabetic or hypertension status, the co-occurrence of retinopathy and CKD or CVD further increased the all-cause mortality compared to those without either condition. CONCLUSIONS: The co-occurrence of retinopathy and systemic vascular conditions predicted a further increase in the risk of mortality. More extensive vascular risk factor assessment and management are needed to detect the burden of vascular pathologies and improve long-term survival in individuals with retinopathy.
  • Item
    Thumbnail Image
    Association of visual impairment with risk for future Parkinson's disease
    Zhu, Z ; Hu, W ; Liao, H ; Tan, Z ; Chen, Y ; Shi, D ; Shang, X ; Zhang, X ; Huang, Y ; Yu, H ; Wang, W ; He, M ; Yang, X (ELSEVIER, 2021-12)
    BACKGROUND: Although visual dysfunction is one of the most common non-motor symptoms among patients with Parkinson's disease (PD), it is not known whether visual impairment (VI) predates the onset of clinical PD. Therefore, we aim to examine the association of VI with the future development of PD in the UK Biobank Study. METHODS: The UK Biobank Study is one of the largest cohort studies of health, enrolling over 500,000 participants aged 40-69 years between 2006 and 2010 across the UK. VI was defined as a habitual distance visual acuity (VA) worse than 0·3 logarithm of the minimum angle of resolution (LogMAR) in the better-seeing eye. Incident cases of PD were determined by self report data, hospital admission records or death records, whichever came first. Multivariable Cox proportional hazard regression models were used to investigate the association between VI and the risk of incident PD. FINDINGS: A total of 117,050 participants were free of PD at the baseline assessment. During the median observation period of 5·96 (IQR: 5·77-6·23) years, PD occurred in 222 (0·19%) participants. Visually impaired participants were at a higher risk of developing PD than non-VI participants (p < 0·001). Compared with the non-VI group, the adjusted hazard ratio was 2·28 (95% CI 1·29-4·05, p = 0·005) in the VI group. These results were consistent in the sensitivity analysis, where incident PD cases diagnosed within one year after the baseline assessment were excluded. INTERPRETATION: This cohort study found that VI was associated with an increased risk of incident PD, suggesting that VI may serve as a modifiable risk factor for prevention of future PD.
  • Item
    Thumbnail Image
    Findings from the 45 and Up Study: smoking is not associated with the risk of early-onset cataract
    Zhang, J ; Han, X ; Wang, W ; Ha, J ; Liu, Z ; Shang, X ; Zhang, L ; Tan, X ; He, M ; Luo, L (AME PUBL CO, 2021-07)
    BACKGROUND: To determine if tobacco smoking is a risk factor for early-onset cataracts. METHODS: This was a prospective population-based cohort study. A total of 70,886 participants aged 45-55 years in the 45 and Up Study were included in our analysis. Early-onset cataracts (EOC) were defined as cataract surgeries performed before 65 years old, based on participant data linked to the Medicare Benefits Schedule (MBS). Smoking habits were assessed at baseline, based on a self-administered questionnaire. A Cox proportional hazards model was used to evaluate the association between tobacco smoking and the risk of early-onset cataracts over the follow-up period. RESULTS: At baseline recruitment, 59.9% of study participants never smoked, 30.5% were former smokers, and 9.6% were current smokers. A total of 1,713 participants underwent cataract surgery over a mean follow-up of 625,042 person-years, with an incidence of 2.74 cases per 1,000 person-years. For current smokers, patients with EOC had longer smoking durations (P=0.019). For former smokers, patients with EOC had higher smoking intensities (P=0.001), were older at smoking commencement (P=0.011), and longer times since quitting (P=0.04). The risk of EOC was not found to be significantly different between current smokers or former smokers, compared to those who had never smoked. Both stratification and sensitivity analyses by gender, surgery year, alcohol intake, physical activity, and income yielded similar results. CONCLUSIONS: Smoking has neither a beneficial nor harmful effect on the long-term incidence of EOC.
  • Item
    Thumbnail Image
    Real-world assessment of topical glaucoma medication persistence rates based on national pharmaceutical claim data in a defined population
    Zhu, Z ; Jiang, Y ; Wang, W ; Scheetz, J ; Shang, X ; Zhang, L ; He, M (WILEY, 2019-09)
    IMPORTANCE: The rate and determinants of persistence to topical glaucoma medications are important for identifying patients at high risk of discontinuing medications and designing targeted approaches to improve persistence. BACKGROUND: To evaluate the rate and determinants of persistence to topical glaucoma medications among middle-aged and older Australian adults. DESIGN: Population-based cohort study. PARTICIPANTS: Participants in need of persistent topical glaucoma medications in the 45 and Up Study. METHODS: The 45 and Up Study is a large-scale population-based cohort study. Participants were classified as needing persistent topical glaucoma medications if at least three claims with related prescriptions were recorded. Persistence was defined as topical glaucoma medications were filled within 90 days. MAIN OUTCOME MEASURES: The rates and determinants of medication persistence at 2-year follow-up. RESULTS: A total of 12 899 patients requiring persistent topical glaucoma medications were identified. Among them, 9019 (69.9%) had persisted with their glaucoma medications for at least 2 years. Multiple logistic regression analysis documented significant effects of patient-related factors (gender, socioeconomic status, language spoken at home, lifestyle and comorbidities) and drug-related factors (total number and drug class) on the persistence rate. Those most at risk groups of non-persistence were those patients living in remote areas (odds ratio, OR: 0.59, 95% confidence interval, CI: 0.37-0.92), having family income over 70 000 AUD/year (OR: 0.53, 95% CI: 0.45-0.62), speaking other languages at home (OR: 0.61, 95% CI: 0.53-0.68), and using cholinergic classes of medications (OR: 0.55, 95% CI: 0.38-0.79). CONCLUSIONS AND RELEVANCE: Our data has shown a medium level of persistence to topical glaucoma medication among middle-aged and older Australian adults. However, efforts are still needed to improve the rate of persistence.
  • Item
    Thumbnail Image
    Predicting the Development of Type 2 Diabetes in a Large Australian Cohort Using Machine-Learning Techniques: Longitudinal Survey Study
    Zhang, L ; Shang, X ; Sreedharan, S ; Yan, X ; Liu, J ; Keel, S ; Wu, J ; Peng, W ; He, M (JMIR PUBLICATIONS, INC, 2020-07)
    BACKGROUND: Previous conventional models for the prediction of diabetes could be updated by incorporating the increasing amount of health data available and new risk prediction methodology. OBJECTIVE: We aimed to develop a substantially improved diabetes risk prediction model using sophisticated machine-learning algorithms based on a large retrospective population cohort of over 230,000 people who were enrolled in the study during 2006-2017. METHODS: We collected demographic, medical, behavioral, and incidence data for type 2 diabetes mellitus (T2DM) in over 236,684 diabetes-free participants recruited from the 45 and Up Study. We predicted and compared the risk of diabetes onset in these participants at 3, 5, 7, and 10 years based on three machine-learning approaches and the conventional regression model. RESULTS: Overall, 6.05% (14,313/236,684) of the participants developed T2DM during an average 8.8-year follow-up period. The 10-year diabetes incidence in men was 8.30% (8.08%-8.49%), which was significantly higher (odds ratio 1.37, 95% CI 1.32-1.41) than that in women at 6.20% (6.00%-6.40%). The incidence of T2DM was doubled in individuals with obesity (men: 17.78% [17.05%-18.43%]; women: 14.59% [13.99%-15.17%]) compared with that of nonobese individuals. The gradient boosting machine model showed the best performance among the four models (area under the curve of 79% in 3-year prediction and 75% in 10-year prediction). All machine-learning models predicted BMI as the most significant factor contributing to diabetes onset, which explained 12%-50% of the variance in the prediction of diabetes. The model predicted that if BMI in obese and overweight participants could be hypothetically reduced to a healthy range, the 10-year probability of diabetes onset would be significantly reduced from 8.3% to 2.8% (P<.001). CONCLUSIONS: A one-time self-reported survey can accurately predict the risk of diabetes using a machine-learning approach. Achieving a healthy BMI can significantly reduce the risk of developing T2DM.
  • Item
    Thumbnail Image
    Effect of physical activity on reducing the risk of diabetic retinopathy progression: 10-year prospective findings from the 45 and Up Study
    Yan, X ; Han, X ; Wu, C ; Shang, X ; Zhang, L ; He, M ; Oyeyemi, AL (PUBLIC LIBRARY SCIENCE, 2021-01-14)
    OBJECTIVE: To examine the association of physical activities (PA) with diabetic retinopathy (DR) progression based on a 10-year follow-up of a large cohort of working-aged diabetic populations in Australia. METHODS: Nine thousand and eighteen working-aged diabetic patients were enrolled from the baseline of the 45 and Up Study from New South Wales, Australia. Self-reported PA collected by questionnaire at baseline in 2006 was graded into low (<5 sessions/week), medium (≥5-14), and high (≥14) levels. Retinal photocoagulation (RPC) treatment during the follow-up period was used as a surrogate for DR progression and was tracked through the Medicare Benefits Schedule, which was available from 2004 to 2016. Cox regression was used to estimate the association between PA and RPC incidence. RESULTS: In the fully adjusted model, higher PA level was significantly associated with a lower risk of RPC incident (Cox-regression, p-value for trend = 0.002; medium vs. low, hazard ratio (HR) = 0.78, 95% Confidence Interval (CI): 0.61-0.98; high vs. low, HR = 0.61, 95%CI: 0.36-0.84. In addition, gender, body mass index, insulin treatment, family history of diabetes, history of cardiovascular disease were significant effect modifiers for the association between PA and RPC. CONCLUSIONS: Higher PA level was independently associated with a lower risk of DR progression among working-aged diabetic populations in this large cohort study.