Discrimination of indoor versus outdoor environmental state with machine learning algorithms in myopia observational studies

Smartwatch Measures of Outdoor Exposure and Myopia in Children

Importance

Time spent outdoors has been proven effective in preventing myopia, but little is known about the association of outdoor exposure patterns with myopia.

Objective

To examine the association of outdoor exposure patterns with myopic shift in children.

Design, Setting, and Participants

This 1-year prospective cohort study from December 2017 to December 2018 was a secondary analysis of a cluster-randomized trial (Shanghai Time Outside to Reduce Myopia [STORM]). STORM was a school-based intervention study, recruiting 16 schools from 8 districts in Shanghai, from October 2016 to December 2018. Children without myopia at baseline who consistently wore a smartwatch for a minimum of 6 hours daily, sustained for at least 90 days, and who had complete information were included. Data analysis was performed from December 2017 to December 2018.

Exposures

The outdoor exposure pattern was defined as the episode of time outdoors and instant sunlight intensity over a continuous period.

Main Outcomes and Measures

Myopic shift was defined as the absolute change in refraction between the initial spherical equivalence and the follow-up spherical equivalence.

Results

This study included 2976 students (mean [SD] age, 7.2 [0.6] years; 1525 girls [51.2%]). The mean (SD) daily time outdoors was 90 (28) minutes, and the mean (SD) sunlight intensity was 2345 (486) lux. Of the 12 outdoor exposure patterns, the major outdoor exposure patterns were time outdoors with at least 15 minutes, accounting for 74.9% of minutes (33 677 584 of 45 016 800 minutes). Only patterns with at least 15 minutes accompanied with no less than 2000 lux were associated with less myopic shift in refraction (for ≥15 minutes and 2000 to 3999 lux, -0.007 diopter [D] [95% CI, -0.011 to -0.002 D]; for ≥15 minutes and ≥4000 lux, -0.006 D [95% CI, -0.010 to -0.002 D]). The isotemporal substitution of patterns with at least 15 minutes and 2000 lux for other outdoor exposure patterns was positively associated with less myopic shift.

Conclusions and Relevance

In this 1-year prospective cohort study of children with smartwatches, continuous outdoor exposure with at least 15 minutes accompanied with no less than 2000 lux sunlight intensity was associated with less myopic shift. These findings suggest that future outdoor interventions should focus not only on the overall time outdoors but also on the effective outdoor exposure patterns, as a means to effectively prevent myopia in children.

Smart Devices in Optometry: Current and Future Perspectives to Clinical Optometry

There is a huge unmet need for eye care with more than a hundred million people living without basic eye care services and facilities. There is an exigency to deploy adequate resources in terms of manpower and equipment to address this. The usage of smart devices in optometry and eye care practice has been gaining momentum for last half a decade, due to the COVID-19 pandemic and technological advancements in telemedicine. These smart devices will help facilitate remote monitoring of important visual functions, ocular signs and symptoms, thus providing better eye care services and facilities and promoting outreach services. Smart devices in optometry exist in the form of gadgets that can be worn in the wrist, and spectacle-mounted or head-mounted devices. On the other hand, with the ubiquitous nature of smartphones, a large number of smartphone applications have been developed and tested for advanced optometry and primary eye care practice, which may potentially reduce the burden of inadequate resources and the unmet need for eye care. This article aims to give an overview of the current trends and future perspectives on the application of such smart devices in optometric practice.

Development and validation of a smartwatch algorithm for differentiating physical activity intensity in health monitoring

To develop and validate a machine learning based algorithm to estimate physical activity (PA) intensity using the smartwatch with the capacity to record PA and determine outdoor state. Two groups of participants, including 24 adults (13 males) and 18 children (9 boys), completed a sequential activity trial. During each trial, participants wore a smartwatch, and energy expenditure was measured using indirect calorimetry as gold standard. The support vector machine algorithm and the least squares regression model were applied for the metabolic equivalent (MET) estimation using raw data derived from the smartwatch. Exercise intensity was categorized based on MET values into sedentary activity (SED), light activity (LPA), moderate activity (MPA), and vigorous activity (VPA). The classification accuracy was evaluated using area under the ROC curve (AUC). The METs estimation accuracy were assessed via the mean absolute error (MAE), the correlation coefficient, Bland-Altman plots, and intraclass correlation (ICC). A total of 24 adults aged 21-34 years and 18 children aged 9-13 years participated in the study, yielding 1790 and 1246 data points for adults and children respectively for model building and validation. For adults, the AUC for classifying SED, MVPA, and VPA were 0.96, 0.88, and 0.86, respectively. The MAE between true METs and estimated METs was 0.75 METs. The correlation coefficient and ICC were 0.87 (p < 0.001) and 0.89, respectively. For children, comparable levels of accuracy were demonstrated, with the AUC for SED, MVPA, and VPA being 0.98, 0.89, and 0.85, respectively. The MAE between true METs and estimated METs was 0.80 METs. The correlation coefficient and ICC were 0.79 (p < 0.001) and 0.84, respectively. The developed model successfully estimated PA intensity with high accuracy in both adults and children. The application of this model enables independent investigation of PA intensity, facilitating research in health monitoring and potentially in areas such as myopia prevention and control.

Effect modification of time spent outdoors on the association between early childhood overweight and myopia: a one-year follow-up study

BACKGROUND

This study examined the moderating role of outdoor time on the relationship between overweight and myopia.

METHODS

The data for this study was obtained from a prospective study in Shanghai, where non-myopic children wore wristwear and were followed up for 1 year. Eye examinations were performed at each visit. The modification effect was assessed on the additive scale using multivariable logistic regression, and relative excess risk due to interaction was used to calculate the modification effect.

RESULTS

A total of 4683 non-myopic children were included with 32.20% being overweight at baseline. Following a 1-year period, 17.42% of children had myopia. When compared to those who spent <90 minutes outdoors, children who spent >120 had a relative risk of myopia onset that was reduced to 0.61. As time spent outdoors decreased, more risks of myopia onset were identified among overweight children than among normal children, the modification effect on the additive scale was -0.007, with ~70% of this effect attributed to the modifying influence of outdoor time.

CONCLUSIONS

Increasing outdoor time can reduce myopia more among overweight children than normal. Future interventions should focus on outdoor activities among overweight children to reduce myopia risks.

Insights into artificial intelligence in myopia management: from a data perspective

Given the high incidence and prevalence of myopia, the current healthcare system is struggling to handle the task of myopia management, which is worsened by home quarantine during the ongoing COVID-19 pandemic. The utilization of artificial intelligence (AI) in ophthalmology is thriving, yet not enough in myopia. AI can serve as a solution for the myopia pandemic, with application potential in early identification, risk stratification, progression prediction, and timely intervention. The datasets used for developing AI models are the foundation and determine the upper limit of performance. Data generated from clinical practice in managing myopia can be categorized into clinical data and imaging data, and different AI methods can be used for analysis. In this review, we comprehensively review the current application status of AI in myopia with an emphasis on data modalities used for developing AI models. We propose that establishing large public datasets with high quality, enhancing the model's capability of handling multimodal input, and exploring novel data modalities could be of great significance for the further application of AI for myopia.

Using Light Meters to Investigate the Light-Myopia Association - A Literature Review of Devices and Research Methods

With the increasing prevalence of myopia, evaluating its relationship with objective light exposure as a potential adjustable environmental factor in myopia development has been an emerging research field in recent years. From a thorough literature search, we identify ten wearable light meters from human studies on light exposure and myopia and present an overview of their parameters, thereby demonstrating the wide between-device variability and discussing its implications. We further identify 20 publications, including two reanalyses, reporting investigations of light-myopia associations with data from human subjects wearing light meters. We thoroughly review the publications with respect to general characteristics, aspects of data collection, participant population, as well as data analysis and interpretation, and also assess potential patterns regarding the absence or presence of light-myopia associations in their results. In doing so, we highlight areas in which more research is needed as well as several aspects that warrant consideration in the study of light exposure and myopia.

Integrated multi-omics and machine learning approach reveals lipid metabolic biomarkers and signaling in age-related meibomian gland dysfunction

Meibomian gland dysfunction (MGD) is a prevalent inflammatory disorder of the ocular surface that significantly impacts patients' vision and quality of life. The underlying mechanism of aging and MGD remains largely uncharacterized. The aim of this work is to investigate lipid metabolic alterations in age-related MGD (ARMGD) through integrated proteomics, lipidomics and machine learning (ML) approach. For this purpose, we collected samples of female mouse meibomian glands (MGs) dissected from eyelids at age two months (n = 9) and two years (n = 9) for proteomic and lipidomic profilings using the liquid chromatography with tandem mass spectrometry (LC-MS/MS) method. To further identify ARMGD-related lipid biomarkers, ML model was established using the least absolute shrinkage and selection operator (LASSO) algorithm. For proteomic profiling, 375 differentially expressed proteins were detected. Functional analyses indicated the leading role of cholesterol biosynthesis in the aging process of MGs. Several proteins were proposed as potential biomarkers, including lanosterol synthase (Lss), 24-dehydrocholesterol reductase (Dhcr24), and farnesyl diphosphate farnesyl transferase 1 (Fdft1). Concomitantly, lipidomic analysis unveiled 47 lipid species that were differentially expressed and clustered into four classes. The most notable age-related alterations involved a decline in cholesteryl esters (ChE) levels and an increase in triradylglycerols (TG) levels, accompanied by significant differences in their lipid unsaturation patterns. Through ML construction, it was confirmed that ChE(26:0), ChE(26:1), and ChE(30:1) represent the most promising diagnostic molecules. The present study identified essential proteins, lipids, and signaling pathways in age-related MGD (ARMGD), providing a reference landscape to facilitate novel strategies for the disease transformation.

Time outdoors positively associates with academic performance: a school-based study with objective monitoring of outdoor time

BACKGROUND

To explore the relationship between outdoor time and academic performance among school-aged children.

METHODS

This study was designed as a cross-sectional study. Data were derived from a school-based prospective children myopia intervention study (STORM). Outdoor time was recorded by self-developed algorithm-validated wristwatches in real-time and calculated as the cumulative average of 10 months. The academic performance was recorded and provided by the participating schools and further standardized. Other information was collected using an online standardized questionnaire. Mixed-effects model and B-Spline method were used to investigate the association between time spent on different types of daily activity, including outdoor activity and academic performance.

RESULTS

A total of 3291 children with mean age 9.25 years were included in the final analysis. Overall, outdoor time was associated with academic performance in a non-linear manner; specifically, not exceeding 2.3 h per day, outdoor time was positively associated with academic performance; exceeding 2.3 h per day, this association became non-significant. Likewise, daily sleep duration and out-of-school learning time were associated with academic performance in a non-linear manner, resulting in turning points of 11.3 and 1.4 h per day, respectively. Separate analysis showed that outdoor time and sleep duration but not out-of-school learning time were positively associated with academic performance in Chinese, mathematics and English.

CONCLUSION

Outdoor time, sleep duration and out-of-school learning time were associated with academic performance in a non-linear manner. Promotion of outdoor time may not negatively impact on academic performance.

TRIAL REGISTRATION

Our study was registered in ClinicalTrials.gov (Identifier: NCT02980445).

Uncovering the role of transient receptor potential channels in pterygium: a machine learning approach

OBJECTIVES

We aimed at identifying the role of transient receptor potential (TRP) channels in pterygium.

METHODS

Based on microarray data GSE83627 and GSE2513, differentially expressed genes (DEGs) were screened and 20 hub genes were selected. After gene correlation analysis, 5 TRP-related genes were obtained and functional analyses of gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) were performed. Multifactor regulatory network including mRNA, microRNAs (miRNAs) and transcription factors (TFs) was constructed. The 5 gene TRP signature for pterygium was validated by multiple machine learning (ML) programs including support vector classifiers (SVC), random forest (RF), and k-nearest neighbors (KNN). Additionally, we outlined the immune microenvironment and analyzed the candidate drugs. Finally, in vitro experiments were performed using human conjunctival epithelial cells (CjECs) to confirm the bioinformatics results.

RESULTS

Five TRP-related genes (MCOLN1, MCOLN3, TRPM3, TRPM6, and TRPM8) were validated by ML algorithms. Functional analyses revealed the participation of lysosome and TRP-regulated inflammatory pathways. A comprehensive immune infiltration landscape and TFs-miRNAs-mRNAs network was studied, which indicated several therapeutic targets (LEF1 and hsa-miR-455-3p). Through correlation analysis, MCOLN3 was proposed as the most promising immune-related biomarker. In vitro experiments further verified the reliability of our in silico results and demonstrated that the 5 TRP-related genes could influence the proliferation and proinflammatory signaling in conjunctival tissue contributing to the pathogenesis of pterygium.

CONCLUSIONS

Our study suggested that TRP channels played an essential role in the pathogenesis of pterygium. The identified pivotal biomarkers (especially MCOLN3) and pathways provide novel directions for future mechanistic and therapeutic studies for pterygium.

Time Outdoors in Reducing Myopia: A School-Based Cluster Randomized Trial with Objective Monitoring of Outdoor time and Light Intensity

PURPOSE

To evaluate the efficacy of time outdoors per school day over 2 years on myopia onset and shift.

DESIGN

A prospective, cluster-randomized, examiner-masked, three-arm trial.

PARTICIPANTS

A total of 6295 students aged 6 to 9 years from 24 primary schools in Shanghai, China, stratified and randomized by school in a 1:1:1 ratio to control (n=2037), test I (n=2329), or test II (n=1929) group.

METHODS

An additional 40 or 80-minutes of outdoor time was allocated to each school day for test I and II groups. Children in the control group continued their habitual outdoor time. Objective monitoring of outdoor and indoor time and light intensity each day was measured with a wrist-worn wearable during the second-year follow-up.

MAIN OUTCOME MEASURES

The 2-year cumulative incidence of myopia (defined as cycloplegic spherical equivalent [SE] of ≤-0.5 diopters[D] at the right eye) among the students without myopia at baseline and changes in SE and axial length (AL) after 2 years.

RESULTS

The unadjusted 2-year cumulative incidence of myopia was 24.9%, 20.6%, and 23.8% for control, test I, and II groups. The adjusted incidence decreased by 16% [Incidence Risk Ratio (IRR)=0.84, 95%CI: 0.72∼0.99; P =0.035] in test I and 11% (IRR=0.89, 95%CI: 0.79∼0.99; P =0.041) in test II when compared with the control group. The test groups showed less myopic shift and axial elongation compared with the control group (test I: -0.84D and 0.55mm, test II: -0.91D and 0.57mm, control: -1.04D and 0.65mm). There was no significant difference in the adjusted incidence of myopia and myopic shift between the two test groups. The test groups had similar outdoor time and light intensity (test I: 127±30 minutes/day and 3557±970 lux/minute; test II: 127±26 minutes/day and 3662±803 lux/minute), but significantly more outdoor time and higher light intensity compared with the control group (106±27 minutes/day and 2984±806 lux/minute). Daily outdoor time of 120∼150 minutes at 5000 lux/minutes or cumulative outdoor light intensity of 600,000∼750,000 lux significantly reduced the IRR by 17%∼31%.

CONCLUSIONS

Increasing outdoor time reduced the risk of myopia onset and myopic shifts, especially in nonmyopic children. The protective effect of outdoor time was related to the duration of exposure as well as light intensity. The dose-response effect between test I and test II was not observed probably due to insufficient outdoor time achieved in the test groups, which suggests that proper monitoring on the compliance on outdoor intervention is critical if one wants to see the protective effect.

Novel Uses and Challenges of Artificial Intelligence in Diagnosing and Managing Eyes with High Myopia and Pathologic Myopia

Myopia is a global health issue, and the prevalence of high myopia has increased significantly in the past five to six decades. The high incidence of myopia and its vision-threatening course emphasize the need for automated methods to screen for high myopia and its serious form, named pathologic myopia (PM). Artificial intelligence (AI)-based applications have been extensively applied in medicine, and these applications have focused on analyzing ophthalmic images to diagnose the disease and to determine prognosis from these images. However, unlike diseases that mainly show pathologic changes in the fundus, high myopia and PM generate even more data because both the ophthalmic information and morphological changes in the retina and choroid need to be analyzed. In this review, we present how AI techniques have been used to diagnose and manage high myopia, PM, and other ocular diseases and discuss the current capacity of AI in assisting in preventing high myopia.

Temporal trends in frequency, type and severity of myopia and associations with key environmental risk factors in the UK: Findings from the UK Biobank Study

This study investigated temporal trends in the epidemiology of primary myopia and associations with key environmental risk factors in a UK population. Data were collected at recruitment (non-cycloplegic autorefraction, year of birth, sex, ethnicity, highest educational attainment, reason and age of first wearing glasses and history of eye disease) from 107,442 UK Biobank study participants aged 40 to 69 years, born between 1939 and 1970. Myopia was defined as mean spherical equivalent (MSE) ≤-1 dioptre (D). Temporal changes in myopia frequency by birth cohort (5-year bands using date of birth) and associations with environmental factors were analysed, distinguishing both type (childhood-onset, <18 years versus adult-onset) and severity (three categories: low -1.00 to -2.99D, moderate -3.00 to -5.99D or high ≥-6.00D). Overall myopia frequency increased from 20.0% in the oldest cohort (births 1939-1944) to 29.2% in the youngest (1965-1970), reflecting a relatively higher increase in frequency of adult-onset and low myopia. Childhood-onset myopia peaked in participants born in 1950-54, adult-onset myopia peaked in the cohort born a decade later. The distribution of MSE only shifted for childhood-onset myopia (median: -3.8 [IQR -2.4, -5.4] to -4.4 [IQR -3.0, -6.2]). The magnitude of the association between higher educational attainment (proxy for educational intensity) and myopia overall increased over time (adjusted Odds Ratio (OR) 2.7 [2.5, 2.9] in the oldest versus 4.2 [3.3, 5.2] in the youngest cohort), being substantially greater for childhood-onset myopia (OR 3.3 [2.8, 4.0] to 8.0 [4.2, 13]). Without delineating childhood-onset from adult-onset myopia, important temporal trends would have been obscured. The differential impact of educational experience/intensity on both childhood-onset and high myopia, amplified over time, suggests a cohort effect in gene-environment interaction with potential for increasing myopia frequency if increasing childhood educational intensity is unchecked. However, historical plateauing of myopia frequency does suggest some potential for effective intervention.