Methods: A child-sized mannequin head was developed to measure light illuminance levels with and without sun-protective equipment, across a wide range of environments in Singapore, outdoors (open park, under a tree, street) and indoors (under a fluorescent illumination with window, under white LED-based lighting without window). A comparison was made between indoor and outdoor light levels that are experienced while children are involved in day-to-day activities.

Results: Outdoor light levels were much higher (11,080-18,176 lux) than indoors (112-156 lux). The higher lux levels protective of myopia (>1000 lux) were measured at the tree shade (5556-7876 lux) and with hat (4112-8156 lux). Sunglasses showed lux levels between 1792 and 6800 lux. Although with sunglasses readings were lower than tree shade and hat, light levels were still 11 to 43 times higher than indoors.

Conclusions: Recommendations on spending time outdoors for myopia prevention with adequate sun protection should be provided while partaking in outdoor activities, including protection under shaded areas, wearing a hat or sunglasses, sunscreen, and adequate hydration.

Methods: We included adults with VI from their second visit of the Singapore Epidemiology of Eye Disease Study. Data on eyecare utilization and spectacle affordability were collected. Low eyecare utilization was defined as no eye check ever or eye checks not even once per year in reference to at least once per year. Difficulty affording glasses was defined as glasses being rated as expensive in reference to not expensive.

Results: There were 985 adults (14.5%; 415 Malays, 260 Indian, and 310 Chinese; mean age [SD]: 69.5 [10.2] years; 55.4% women) with VI who answered the above questions, were included. Of these, 624 (63.4%) wore glasses. The rates of low eyecare utilization and difficulty affording eyeglasses were 31% and 63%, respectively. Compared to Chinese (23.8%) and Indians (18.8%), Malays (57.4%) had the highest rates of low eyecare utilization (P < 0.001), and most difficulty affording eyeglasses (47.2% vs. 26.1% and 26.6% in Chinese and Indians, respectively; P < 0.001). Younger age, low socioeconomic status, absence of diabetes, absence of self-reported eye conditions, and poor vision were independently associated with low eyecare utilization, whereas older age and female sex was associated with difficulty affording glasses.

Conclusions: In this multi-ethnic population with VI, almost one-third had low eyecare utilization and nearly two-thirds reported difficulty affording eyeglasses.

METHODS: The development of the FitSight fitness tracker involved the designing of two components: (1) the smartwatch with custom-made FitSight watch application (app) to log the instant light illuminance levels the wearer is exposed to, and (2) a companion smartphone app that synchronizes the time outdoors recorded by the smartwatch to smartphone via Bluetooth communication. Smartwatch wear patterns and tracker-recorded daily light illuminance levels data were gathered over 7 days from 23 Singapore children (mean ± standard deviation age: 9.2 ± 1.4 years). Feedback about the tracker was obtained from 14 parents using a three-level rating scale: very poor/poor/good.

RESULTS: Of the 14 parents, 93% rated the complete "FitSight fitness tracker" as good and 64% rated its wearability as good. While 61% of 23 children wore the watch on all study days (i.e., 0 nonwear days), 26% had 1 nonwear day, and 4.5% children each had 3, 4, and 5 nonwear days, respectively. On average, children spent approximately 1 hour in light levels greater than 1000 lux on weekdays and 1.3 hours on weekends (60 ± 46 vs. 79 ± 53 minutes, P = 0.19). Mean number of outdoor "spurts" (light illuminance levels >1000 lux) per day was 8 ± 3 spurts with spurt duration of 34 ± 32 minutes.

CONCLUSION: The FitSight tracker with its novel features may motivate children to increase time outdoors and play an important role in supplementing community outdoor programs to prevent myopia.

Methods: We collected 3794 corneal images from 542 eyes of 280 subjects and developed seven deep learning models based on anterior and posterior eccentricity, anterior and posterior elevation, anterior and posterior sagittal curvature, and corneal thickness maps to extract deep corneal features. An independent subset with 1050 images collected from 150 eyes of 85 subjects from a separate center was used to validate models. We developed a hybrid deep learning model to detect KCN. We visualized deep features of corneal parameters to assess the quality of learning subjectively and computed area under the receiver operating characteristic curve (AUC), confusion matrices, accuracy, and F1 score to evaluate models objectively.

Results: In the development dataset, 204 eyes were normal, 123 eyes were suspected KCN, and 215 eyes had KCN. In the independent validation dataset, 50 eyes were normal, 50 eyes were suspected KCN, and 50 eyes were KCN. Images were annotated by three corneal specialists. The AUC of the models for the two-class and three-class problems based on the development set were 0.99 and 0.93, respectively.

Conclusions: The hybrid deep learning model achieved high accuracy in identifying KCN based on corneal maps and provided a time-efficient framework with low computational complexity.