Artificial Intelligence-Assisted Prescription Determination for Orthokeratology Lens Fitting: From Algorithm to Clinical Practice.

OBJECTIVES: To explore the potential of artificial intelligence (AI) to assist prescription determination for orthokeratology (OK) lenses. METHODS: Artificial intelligence algorithm development followed by a real-world trial. A total of 11,502 OK lenses fitting records collected from seven clinical environments covering major brands. Records were randomly divided in a three-way data split. Cross-validation was used to identify the most accurate algorithm, followed by an evaluation using an independent test data set. An online AI-assisted system was implemented and assessed in a real-world trial involving four junior and three senior clinicians. RESULTS: The primary outcome measure was the algorithm's accuracy (ACC). The ACC of the best performance of algorithms to predict the targeted reduction amplitude, lens diameter, and alignment curve of the prescription was 0.80, 0.82, and 0.83, respectively. With the assistance of the AI system, the number of trials required to determine the final prescription significantly decreased for six of the seven participating clinicians (all P <0.01). This reduction was more significant among junior clinicians compared with consultants (0.76±0.60 vs. 0.32±0.60, P <0.001). Junior clinicians achieved clinical outcomes comparable to their seniors, as 93.96% (140/149) and 94.44% (119/126), respectively, of the eyes fitted achieved unaided visual acuity no worse than 0.8 ( P =0.864). CONCLUSIONS: AI can improve prescription efficiency and reduce discrepancies in clinical outcomes among clinicians with differing levels of experience. Embedment of AI in practice should ultimately help lessen the medical burden and improve service quality for myopia boom emerging worldwide.

Efficacy of contact lenses for myopia control: Insights from a randomised, contralateral study design.

PURPOSE: To determine the efficacy of two myopia control contact lenses (CL) compared with a single-vision (SV) CL. METHODS: Ninety-five Chinese children with myopia, aged 7-13 years in a 1-year prospective, randomised, contralateral, cross-over clinical trial with 3 groups; bilateral SVCL (Group I); randomised, contralateral wear of an extended depth of focus (EDOF) CL and SVCL (Group II) and MiSight® CL and SVCL (Group III). In Groups II and III, CL were crossed over at the 6-month point (Stage 1) and worn for a further 6 months (Stage 2). Group I wore SVCL during both stages. At baseline and the end of each stage, cycloplegic spherical equivalent refractive error (SE) and axial length (AL) were measured. Six-monthly ΔSE/ΔAL across groups was analysed using a linear mixed model (CL type, stage, eye and eye* stage included as factors). Intra-group paired differences between eyes were determined. RESULTS: In Group I, mean (SD) ΔSE/ΔAL with SVCL was -0.41 (0.28) D/0.13 (0.09) mm and -0.25 (0.27) D/0.16 (0.09) mm for stages 1 and 2, with a mean paired difference between eyes of 0.01 D/0.01 mm and 0.05 D/-0.01 mm, respectively. ΔSE/ΔAL with SVCL was similar across Groups I to III (Stage 1: p = 0.89/0.44, Stage 2: p = 0.70/ 0.64). In Groups II and III, ΔSE/ΔAL was lower with the EDOF and MiSight® CL than the contralateral SVCL in 68% to 94% of participants, and adjusted 6-month ΔSE/ΔAL with EDOF was similar to MiSight® (p = 0.49/0.56 for ΔSE/ΔAL, respectively). Discontinuations across the three groups were high, but not different between the groups (33.3%, 48.4% and 50% for Groups I to III, respectively [p = 0.19]) and most discontinuations occurred immediately after baseline. CONCLUSIONS: Extended depth of focus and MiSight® CL demonstrated similar efficacy in slowing myopia. When switched from a myopia control CL to SVCL, myopia progression was similar to that observed with age-matched wearers in SVCL and not suggestive of rebound.

Genipin inhibits the scleral expression of miR-29 and MMP2 and promotes COL1A1 expression in myopic eyes of guinea pigs.

PURPOSE: High myopia can lead to blindness. Genipin is a collagen cross-linking agent that may be used to treat myopia. However, the mechanism of action of genipin for the treatment of myopia is unclear. This study investigated the effect of genipin on the scleral expression of the miR-29 cluster, matrix metalloproteinase 2 (MMP2), and collagen alpha1 chain of type I (COL1A1) in a guinea pig model of myopia. METHODS: The model of myopia was established by treating guinea pigs with a - 8D lens on both eyes for 21 days, and eyes with a refractive error of - 6D or greater were included. Quantitative reverse transcription-polymerase chain reaction (RT-PCR) and western blot were used to examine the mRNA and protein expression, respectively. A dual-luciferase assay was used to determine the direct targeting of the miR-29 cluster on the 3'-untranslated region (UTR) of the COL1A1 gene. RESULTS: The scleral expression of miR-29a, miR-29b, and miR-29c as well as MMP2 was significantly increased, and the scleral expression of COL1A1 was significantly decreased in the myopia group. Genipin treatment reversed these effects in myopic eyes. The dual-luciferase assay showed that the luciferase activities were significantly decreased in human embryonic kidney (HEK) cells transfected with miR-29a and miR-29b, but not miR-29c, compared with those transfected with control miRNAs. CONCLUSIONS: Genipin inhibits the scleral expression of the miR-29 cluster and MMP2 and promotes COL1A1 expression in a guinea pig model of myopia. Thus, genipin may promote COL1A1 expression by reducing the expression of the miR-29 cluster.

Effects of the long wavelength-filtered continuous spectrum on natural refractive development in juvenile guinea pigs.

AIM: To investigate the effects of spectral composition and light intensity on natural refractive development in guinea pigs. METHODS: A total of 124 pigmented guinea pigs (2-week-old) were randomly assigned to three groups at high (Hi; 4000 lx), medium (Me; 400 lx) and low (Lo; 50 lx) light intensities under a 12:12 light/dark cycle for 6wk. Each group was subdivided into subgroups with the following spectra: broad spectrum Solux halogen light (BS), 600 nm above-filtered continuous spectrum (600F), 530 nm above-filtered continuous spectrum (530F), and 480 nm above-filtered continuous spectrum (480F; HiBS: n=10, Hi600F: n=10, Hi530F: n=10, Hi480F: n=10, MeBS: n=10, Me600F: n=10, Me530F: n=10, Me480F: n=10, LoBS: n=11, Lo600F: n=12, Lo530F: n=10, Lo480F: n=11). Refractive error, corneal curvature radius, and axial dimensions were determined by cycloplegic retinoscopy, photokeratometry, and A-scan ultrasonography before and after 2, 4, and 6wk of treatment. Average changes from both eyes in the ocular parameters and refractive error were compared among different subgroups. RESULTS: After 6wk of exposure, high-intensity lighting enhanced hyperopic shift; medium- and low-intensity lighting enhanced myopic shift (P<0.05). Under the same spectrum, axial increase was larger in the low light intensity group than in the medium and high light intensity groups (HiBS: 0.65±0.02 mm, MeBS: 0.67±0.01 mm, LoBS: 0.82±0.02 mm; Hi600F: 0.64±0.02 mm, Me600F: 0.67±0.01 mm, Lo600F: 0.81±0.01 mm; Hi530F: 0.64±0.02 mm, Me530F: 0.67±0.01 mm, Lo530F: 0.73±0.02 mm; Hi480F: 0.64±0.01 mm, Me480F: 0.66±0.01 mm, Lo480F: 0.72±0.02 mm; P<0.05). Under 400 lx, there was a faster axial increase in the MeBS group than in the Me480F group (P<0.05). Under 50 lx, axial length changes were significantly larger in LoBS and Lo600F than in Lo530F and Lo480F (P<0.01). CONCLUSION: Under high-intensity lighting, high light intensity rather than spectrum distributions that inhibits axial increase. Under medium- and low-intensity lighting, filtering out the long wavelength inhibits axial growth in juvenile guinea pigs.

A novel apparatus for interocular interaction evaluation in children with and without anisometropic amblyopia.

BACKGROUND: Dichoptic visual stimulation may be achieved using shutter goggles and mirror systems. These methods vary in their feasibility for use in children. This study aims to investigate the feasibility of use of a simple trial frame-based system to evaluate interactions in children. METHODS: Low contrast acuity, contrast sensitivity and alignment sensitivity were measured in the non-dominant eye of 10 normally-sighted children, 14 anisometropic children without amblyopia and 14 anisometropic amblyopic children (aged 5-11 years) using goggles and a trial frame apparatus (TFA). The dominant eye was either fully or partially occluded. The difference in visual functions in the non-dominant eye between the full and partial occlusion conditions was termed the 'interaction index'. Agreement between the TFA and goggles in terms of visual functions and interactions was assessed in anisometropic children with and without amblyopia using the Bland-Altman method and t-test. Training sessions allowed subjects to become accustomed to the systems and tasks. The duration of training, the number of breaks requested by subjects and their willingness to attend further experiments were recorded in 10 subjects from each group and were compared between groups and between systems. RESULTS: Both Bland-Altman and t-test methods indicated acceptable agreement between the TFA and goggles in visual function and interaction measures (p > 0.05), except for contrast sensitivity measured in anisometropic children without amblyopia (p = 0.042). For all subject groups, contrast sensitivity training was significantly longer using goggles than using the TFA (p ≤ 0.001). Significantly more breaks were requested in acuity and contrast sensitivity testing, when goggles were used than when the TFA was used (p < 0.045). Anisometropic children without amblyopia showed a significantly greater willingness to attend more experiments using the TFA than using goggles (p = 0.025). CONCLUSION: The TFA may be a useful tool in studies of interactions in amblyopes, particularly in studies of children's vision.

[The strategy of determining the customized near addition lens for juvenile esophoric myopes].

OBJECTIVE: To discuss accommodative lag and near phoria varying with the near addition lenses and establish the mathematic equations of near adds and accommodative lag, near phoria respectively in order to getting the optimal adds, which would create the least error in accommodative and convergence responses. METHODS: To measure accommodative lag by Shin-Nippon and near phoria by Cover Test in 33 cm after providing 0D, +0.75D, +1.50D, +2.00D, +2.50D five near adds orderly to myopes after 110 simple myopes with esophoria have been corrected far vision wholly basing on MPMVA norm; to analyze the relation of near adds and accommodative lag, near phoria by linearity regression. RESULTS: While increasing dioptric value of the near adds, accommodative lag and esophoria decrease gradually. Both accommodative lag [F(262.76, 4), P < 0.01]and near phoria [F(223.45, 4), P < 0.01] show statistically significant difference among the five near adds. There is linearity correlation between the optimal adds and initial accommodative lag (r = 0.526, P < 0.01) and between the optimal adds and initial near phoria (r = 0.523, P < 0.01). CONCLUSION: The range of optimal near adds can be gotten basing on initial accommodative lag and near phoria of myopic individual, so that it can be provided referenced strategy of determining the customized near addition lens for juvenile esophoric myopes.