Test Lens Research Articles

Comfort advantages demonstrated with a novel soft contact lens: A randomized clinical trial

PurposeTo compare comfort outcomes between a novel daily disposable contact lens – designed to maximize comfort – and an established control. The hypothesis was that the test lens would be superior to the control for four key comfort questionnaire items: end-of-day comfort, all-day comfort, visual comfort while driving at night, and reduction of ocular fatigue from digital device use. MethodsThis randomized, controlled, subject-masked, parallel-arm study enrolled young (18–39 years), healthy, myopic, contact lens wearers with an up-to-date prescription at 19 investigational sites in the United States. Subjects wore either the test (ACUVUE® OASYS MAX 1-Day, senofilcon A) or control (Dailies Total1®, delefilcon A) lens for 2 weeks of bilateral, daily disposable wear before completing comfort questionnaire items, each of which had 5 or 6 response options. For each item, the odds ratio for positive (top-two-box) responses was estimated from a binomial generalized linear mixed model. A gatekeeping approach combined with the truncated Hochberg procedure was used for multiplicity adjustment. ResultsOf 344 enrolled subjects, 342 subjects were randomized and dispensed lenses, with 171 subjects per lens group. Among the 342 subjects, 68.4% were female, 83.6% were White, and the average age was 29.7 (±5.53) years. The test lens was statistically superior to the control for all four comfort questions: odds ratios (test vs. control) were 2.01 (95% CI: 1.25, 3.22) for end-of-day comfort, 2.17 (alpha-adjusted CI: 1.30, 3.64) for all-day comfort, 2.00 (alpha-adjusted CI: 1.18, 3.41) for reducing ocular fatigue from digital device use, and 1.77 (95% CI: 1.04, 3.02) for comfortable vision while driving at night. ConclusionThe test lens demonstrated statistically superior physical and visual comfort, as measured by the four comfort endpoints, compared to the control. The test lens had significantly greater odds of favorable responses for all comfort items compared to the control.

Single-detector multiplex imaging flow cytometry for cancer cell classification with deep learning.

Imaging flow cytometry, which combines the advantages of flow cytometry and microscopy, has emerged as a powerful tool for cell analysis in various biomedical fields such as cancer detection. In this study, we develop multiplex imaging flow cytometry (mIFC) by employing a spatial wavelength division multiplexing technique. Our mIFC can simultaneously obtain brightfield and multi-color fluorescence images of individual cells in flow, which are excited by a metal halide lamp and measured by a single detector. Statistical analysis results of multiplex imaging experiments with resolution test lens, magnification test lens, and fluorescent microspheres validate the operation of the mIFC with good imaging channel consistency and micron-scale differentiation capabilities. A deep learning method is designed for multiplex image processing that consists of three deep learning networks (U-net, very deep super resolution, and visual geometry group 19). It is demonstrated that the cluster of differentiation 24 (CD24) imaging channel is more sensitive than the brightfield, nucleus, or cancer antigen 125 (CA125) imaging channel in classifying the three types of ovarian cell lines (IOSE80 normal cell, A2780, and OVCAR3 cancer cells). An average accuracy rate of 97.1% is achieved for the classification of these three types of cells by deep learning analysis when all four imaging channels are considered. Our single-detector mIFC is promising for the development of future imaging flow cytometers and for the automatic single-cell analysis with deep learning in various biomedical fields.

Peripheral myopic defocus signal affects the efficiency of visual information processing in myopic children.

Spectacle lenses with peripheral lenslets have shown promise for myopia control by providing peripheral myopic defocus signals. Here, we aimed to investigate the impact of prolonged exposure (>6 months) to peripheral myopic defocus on visual information processing in myopic children. The study included 30 myopic children who habitually wore spectacle lenses with highly aspherical lenslets (HAL group) and 34 children who habitually wore single-vision (SV group) spectacles. The quick contrast sensitivity function (qCSF) was used to measure contrast sensitivity (CS) under conditions of no or high noise. Both groups were tested with HAL and SV lenses. The perceptual template model was utilised to fit the contrast sensitivity function (CSF) and determine differences in information processing efficiency through internal additive noise ( ) and perceptual template gain (β). The areas under the log CSF in the SV group were significantly higher than for the HAL group in both zero-noise conditions with the SV test lens (p = 0.03) and high-noise conditions with the HAL test lens (p = 0.02). For 2 cycle per degree (cpd) stimuli, β was significantly higher in the SV group with the HAL test lens than in the HAL group (p = 0.02), while there was a trend towards a significant difference in β for 6 cpd stimuli (p = 0.07). However, there were no significant differences in between the two groups, with or without noise interference. The reduced CS observed in myopic children wearing HAL lenses for 6 months or more may be due to decreased β. This suggests that prolonged use of spectacle lenses with peripheral myopic defocus signals may compromise the central visual system's ability to process additional external noise, resulting in decreased efficiency in visual information processing.

Quality of vision clinical outcomes for a new fully-refractive extended depth of focus Intraocular Lens

Background/objectiveTo evaluate the visual performance of a purely refractive extended depth of focus (EDF) intraocular lens (IOL).Subjects/methodsA prospective, multi-center, randomized, subject/evaluator-masked study. Subjects were bilaterally implanted with the EDF test (Model ZEN00V, TECNIS PureSee™ IOL, N = 60) or an enhanced monofocal control (Model ICB00, TECNIS Eyhance™ IOL, N = 57) IOL. Monocular corrected distance (CDVA), intermediate (DCIVA), near acuities (DCNVA) and patient reported visual symptoms were evaluated at the 6-month visit. Monocular mesopic contrast sensitivity (CS) and depth of focus (DOF) testing were assessed at 3 months.ResultsCDVA (Mean ± SD) was −0.06 ± 0.08 for test and −0.05 ± 0.08 logMAR for control groups. DCIVA was 0.13 ± 0.08 for test and 0.18 ± 0.14 logMAR for control groups (p = 0.0127). DCNVA was 0.37 ± 0.10 for test and 0.43 ± 0.16 logMAR for control groups (p = 0.0137). Test lens was statistically superior for intermediate and near. Overall, 91.7% (halos), 95.0% (starbursts) and 95.0% (glare) of test lens patients reported that they did not experience, were not bothered, or were slightly bothered by specific visual symptoms, compared to 98.2%, 100% and 96.5% in the control group. The DOF range over which monocular visual acuity was 0.20 logMAR or better was −1.6 D for the test lens. Mesopic CS was comparable between both groups, falling within 0.11 log units for all measured cycles per degree with and without glare.ConclusionThe EDF IOL demonstrated extended range of vision and statistically superior intermediate and near performance compared to the monofocal IOL. Distance visual acuity, contrast sensitivity and dysphotopsia profile were similar to the monofocal IOL.

Focal length measurement based on vortex beam interference

This paper proposes a new method to measure the focal lengths of both convex and concave lenses based on vortex beam interference. Theoretical analysis illustrates that the maximum intensity distribution of the interferogram is in accordance with the Fermat's spiral, and the spiral coefficient is proportional to the difference between the focal length of the test lens and the distance from the test lens to the camera. In order to calculate the precise focal length, we separately block the light paths to obtain the reference and test beam spots, and get the ratio of these two spot sizes. Then, the ratio and the spiral coefficient are used to determine the focal length. For a biconvex lens with a focal length of 200 mm and a biconcave lens with a focal length of −100 mm, the relative measurement errors are 0.0892 % and 0.0800 %, respectively. This research provides a novel and convenient method for measuring the focal length, and no-sophisticated displacement devices and moving the lens to be measured are required in the measurement.

Rotational stability of a new hydrophobic acrylic intraocular lens platform.

To evaluate rotational stability for ease of rotation of a new intraocular lens (IOL) platform with 4 haptics in an ex vivo model and compare it with a control single-piece lens with 2 open loops. Intermountain Ocular Research Center, John A. Moran Eye Center, University of Utah, Salt Lake City, Utah. Experimental study. 10 human cadaver eyes were prepared as per the Miyake-Apple technique. After injection of the test or control lens, clockwise and counterclockwise rotations were attempted with a hook, with and without an ophthalmic viscosurgical device in place. Ease of rotation was scored as (1) very easy, (2) easy, (3) difficult, and (4) very difficult. Rotation of the entire eye containing test or control IOL was also performed with a multipurpose rotator (2 minutes; 220 revolutions per minute) to evaluate its effect on IOL alignment. There were statistically significant differences regarding ease of rotation between test and control lenses in all 4 scenarios, with rotation being more difficult with the test lens ( P < .05, Wilcoxon signed-rank test). No change in the alignment of test or control lenses was observed after eye rotation with the multipurpose rotator. The new IOL platform showed greater rotational stability than the control lens in this model, owing to design features such as 4 small arcs of contact between the haptics and the bag equator, a bulge at the distal end of each haptic, and arcuate haptics with curvatures oriented toward each other.

Anonic Silicon Hydrogels Affect the Concentration of Proteins in Tears during Wear

Purpose The objective of this study was to quantitatively assess the concentration of human tear proteins in patients wearing contact lenses of various ionicities and determine whether differences were related to the incidence of corneal infiltrative events (CIE). Methods 24 subjects (samples) were randomly selected for spectral count analysis to obtain protein concentrations using LCMS analysis. The subjects were neophyte and ametropic with ages between 18 and 40; 6 wore control lenses, 8 wore TestLens1, and 10 wore TestLens2. 16 subjects experienced CIEs during the study. Results A pairwise multiple hypothesis test identified 7 proteins that significantly differed in concentration between TestLens1 and control, and 11 proteins that differed between TestLens2 and control. Of the 12 unique proteins, 9 were at increased concentration and 3 were at lower concentration in the tears of test lens wearers compared to the control lens group. Bootstrap clustering confirmed these findings, showing 3 similar clusters to the original sample groups which separated people wearing control lenses from those wearing TestLens1 or TestLens2 with 83% accuracy and between TestLens1 and TestLens2 with 45% accuracy. Permutation testing identified 5 proteins that had significantly changed in concentration between people wearing TestLens2 and Control lenses. There was no difference in protein concentrations between those subjects who experienced a CIE and those who did not. Conclusion Wearing contact lenses of different ionicities can affect the concentration of proteins in the tear film. The current study did not find any associations of the concentration of proteins with CIEs. Future tests with increased sample size are needed to establish any relations between these changes and clinical performance.

Tear film dynamics of a new soft contact lens.

To present the objective metrics from a study that evaluated the clinical performance of a senofilcon A contact lens, both with and without a new manufacturing technique. This was a single-site, five-visit, controlled, randomised, subject-masked, 2 × 2 crossover study (May-August 2021) with a 2-week lens dispensing period (bilateral wear) and weekly follow-up visits. Healthy adult (18-39 years), habitual spherical silicone hydrogel contact lens wearers were included. The High-definition (HD) Analyzer™ was used to objectively measure the lens-on-eye optical system resulting from the study lenses at 1-week follow-up. Measurements assessed were vision break-up time (VBUT), modulation transfer function (MTF) cutoff, Strehl ratio (SR), potential visual acuity (PVA) for 100% contrast and objective scatter index (OSI). Of the 50 enrolled participants, 47 (94.0%) were randomly assigned to one of the two possible lens wear sequences (test/control or control/test) and dispensed at least one study lens. The estimated odds ratio of VBUT > 10 s was 1.582 (95% confidence interval [CI]: 1.009 to 2.482) in test versus control lens. The least squares mean difference estimates of MTF cutoff, SR and PVA for 100% contrast between test versus control lens were 2.243 (95% CI: 0.012 to 4.475), 0.011 (95% CI: -0.002 to 0.023) and 0.073 (95% CI: -0.001 to 0.147), respectively. The estimated ratio of median OSI between test versus control lens was 0.887 (95% CI: 0.727 to 1.081). The test lens demonstrated superiority over the control lens with respect to VBUT and MTF cutoff. No serious adverse events were reported; eight adverse events (three ocular, five non-ocular) were indicated by six participants during the study. The test lens demonstrated an increased probability of having a longer VBUT (>10 s). Future studies may be designed to assess the efficacy and long-term use of the test lens in a larger population.

Unexpected vision performance with photochromic contact lenses in normal and low light conditions: An analysis of two randomized trials.

Evaluate the performance of a photochromic contact lens in various lighting conditions throughout the day, including those indoor and outdoor environments where the photochromic contact lens is in a less active or inactive state. Data from two clinical trials of a photochromic contact lens were analyzed to evaluate its performance in various light environments. Both studies involved a photochromic test lens (ACUVUE® OASYS with Transitions™ Light Intelligent Technology™) and a similar non-photochromic control lens (ACUVUE® OASYS 2-week with HYDRACLEAR® PLUS). The studies were both multi-visit, multi-site, 2-treatment by 3-period randomized crossover (i.e., Test/Control/Control or Control/Test/Test) dispensing studies, with follow-up visits after each 2-week dispensing period. A total of 250 subjects were dispensed lenses across both studies, of which 237 total subjects completed. In situations where exposure to an activating light source is common (e.g., outdoors), the Test lens was preferred nearly 6:1 over the control lens. In situations where exposure to an activating light source is less common - indoors, driving at night, using digital devices -, the Test lens was still preferred over the control lens by margins of 4:1, nearly 4:1, and over 3:1 respectively. The Test lens was superior with respect to quality of vision, ability to see comfortably, clarity of vision, reduction of squinting while using computers and reduction of bright light while driving at night. The photochromic test contact lens was rated superior to a non-photochromic control lens in environmental situations where the lens is in a less active or inactive state.

Simple and accurate method for determining lens focal length

A simple and accurate method for determining the focal length of a lens is presented. There is no need to know the distance between lens principal planes and can be performed with easily available and inexpensive hardware. Depending on the test lens (TL) f-number, the error on determining focal length can be in the order of 0.1% of the focal length. We discuss determining focal length with the Newton equation f2 = zz ′ , with the Bessel equation f = ( (L − PP)2 − D2 ) / 4 ( L − PP ) , with equation f2 = y ′ ( y + y ′ ) , or with equation f = ( y2 − PP2 ) / 4PP.

Transferability of Diffractive Structure in the Compression Molding of Chalcogenide Glass

This study investigates the use of Ge28Sb12Se60 chalcogenide glass for the compression molding of an infrared optical lens with a diffractive structure. Firstly, a mold core was prepared through ultra-precision grinding of tungsten carbide, and a chalcogenide glass preform was crafted through a polishing process and designed with a radius that would prevent gas isolation during the molding process. The test lens was then molded at various temperature conditions using the prepared mold core and preform. The diffractive structures of both the mold core and the resulting molded lens were analyzed using a microscope and white light interferometer. The comparison of these diffractive structures revealed that the molding temperature had an effect on the transferability of the diffractive structure during the molding of the chalcogenide glass lens. Furthermore, it was determined that, when the molding temperature was properly adjusted, the diffractive structure of the core could be fully transferred to the surface of the chalcogenide lens. Optimized chalcogenide glass-based lenses have the potential to serve as cost-effective yet high-performance IR optics.

The effect of a short-wave filtering contact lens on color appearance

We assessed the effect of a contact lens that filters short-wavelength (SW) visible light on color appearance. These effects were modeled and measured by direct comparison to a clear contact lens. Sixty-one subjects were enrolled, and 58 completed as cohort; 31 were 18 to 39 years old (mean ± SD, 29.6 ± 5.6), 27 were 40 to 65 years old (50.1 ± 8.1). A double-masked contralateral design was used; participants randomly wore a SW-filtering contact lens on one eye and a clear control lens on the other eye. Subjects then mixed three primaries (including a short-wave primary, strongly within the absorbance of the test lens) until a perceived perfect neutral white was achieved with each eye. Color appearance was quantified using chromaticity coordinates measured with a spectral radiometer within a custom-built tricolorimeter. Color vision in natural scenes was simulated using hyperspectral images and cone fundamentals based on a standard observer. Overall, the chromaticity coordinates of matches that were set using the SW-filtering contact lens (n = 58; x = 0.345, y = 0.325, u' = 0.222, v' = 0.470) and clear contact lens (n = 58; x = 0.344, y = 0.325, u' = 0.223, v' = 0.471) were not significantly different, regardless of age group. Simulations indicated that, for natural scenes, the SW-filtering contact lens that was evaluated changes L/(L+M) and S/(L+M) chromatic contrast by no more than -1.4% to +1.1% and -36.9% to +5.0%, respectively. Tricolorimetry was used to measure color appearance in subjects wearing a SW-filtering lens in one eye and a clear lens in the other, and the results indicate that imparting a subtle tint to a contact lens, as in the SW-filtering lens that was evaluated, does not alter color appearance for younger or older subjects. A model of color vision predicted little effect of the lens on chromatic contrast for natural scenes.

Optical calibration system for refractometers

An optical calibration system for refractometers was designed and built using a singlet lens and a translatable detector that induces spherical error, making it a simple and easy-to-use system. The system was calibrated using spectacles and then validated by correcting blurred images created by a series of ophthalmic test lenses. A linear regression model of the correction of the ophthalmic test lenses showed nearly direct correlation with an offset of 0.12 D and an R2 value of 0.9998. An updated version of the system was designed in Zemax OpticStudio with a singlet lens, which could induce spherical error accurate to that of the human eye within ≤ 0.05 D. In future work, this system will be manufactured and tested in the lab.

Spectacle lenses with slightly aspherical lenslets for myopia control: clinical trial design and baseline data

ObjectivesMyopia is a major public health problem and it is essential to find safe and effective means to control its progression. The study design and baseline data are presented for a one-year prospective, double-masked, crossover, randomized clinical trial evaluating the efficacy of single vision spectacle lenses with concentric rings of slightly aspherical contiguous lenslets technology (SAL) on myopia control.MethodsOne hundred 8- to 13-year old Chinese children with a refractive error of -0.75 D to -4.75 D were assigned to two groups. In Group 1, SAL and single vision lenses were each worn for 6 months, and Group 2 wore the lenses in the reversed order. Primary outcomes are axial length and spherical equivalent of cycloplegic refractive error. Secondary outcomes included corneal thickness, anterior chamber depth, lens thickness, visual acuity, and lens adaptation.ResultsNo significant differences in baseline parameters (cycloplegic spherical equivalent, axial length, age) were found between groups (0.49 < p < 0.94). All children adapted well to the test lenses and there was no significant difference in visual acuity between the SAL and single vision lenses (p = 0.27).ConclusionsThe children in the two well balanced groups had comparable visual acuity and adapted well to the test lenses. These results imply that visual acuity can be well improved by SAL lenses. Clear visual acuity provides the assurance for good compliance in this longitudinal study.

Overcoming the error of optical power measurement caused by the curvature radius.

In traditional focimeter measurements, the lens cannot completely coincide with the diaphragm owing to the change of radius, resulting in an increase in the power measurement error with an increase in the lens power. We proposed a method, using the SVM machine learning algorithm, to restore the measurement of the focimeter, using a lens power data set obtained from lens features, obtained through an automatic acquisition system. Total up to 83 groups of single focus lenses with refractive indices of 1.56 and 1.60, ranging from -10 m-1 to + 8 m-1 every 0.25 m-1 and -10.5 m-1 to -15 m-1 every 0.5 m-1, were used for lens image acquisition. The experimental results show that the ten-fold average F1 score of the classification under this method is 100%. The test lens power can be accurately identified and the measurement error can be overcome.

Evaluation of corneal staining with an antihistamine-releasing contact lens

ABSTRACT Clinical relevance Practitioners can be reassured that this antihistamine-releasing contact lens has no additional effect on corneal epithelial integrity. Background To evaluate the effect of an antihistamine-releasing soft contact lens on corneal epithelium integrity when worn on a daily disposable modality for 12 weeks. Methods Two clinical trials using the same randomised, double-masked, placebo-controlled, parallel-group design enrolled healthy contact lens wearers. Participants wore either etafilcon A with 0.019 mg ketotifen (test; n = 374) or etafilcon A with no added drug (placebo; n = 186). Assessments were conducted at baseline, 1 week and 4, 8, and 12 weeks. Slit-lamp evaluations of corneal staining (using sodium fluorescein) in all regions of the corneas of both eyes were graded on a 0–4 scale. Data from all randomised participants were analysed. Results Corneal staining was infrequent and, where present, was mild (Grade 2) or trace (Grade 1). There were no Grade 3 or 4 findings of corneal staining. The overall proportion of findings of Grade 0 corneal staining was 95.86% with the test lens and 95.88% with the placebo lens. The odds of no staining were not statistically different between the test and placebo lenses (Odds Ratio: 0.96, 95% Confidence Intervals: 0.76 to 1.20). There were no serious ocular adverse events or signs of ocular surface medicamentosa. Conclusion Both test and placebo lenses were well tolerated by subjects during the 3 months of wear. The antihistamine-releasing contact lens does not significantly impact corneal epithelial integrity.

Frequently Experienced Problems in the Early Stages after Prescribing Rigid Gas Permeable Contact Lenses

Prescribing rigid gas-permeable (RGP) lenses involves a series of processes that determine the most appropriate final lens through the trial use of test lenses based on the results of slit lamp microscopy, measuring refraction and corneal curvature, and corneal topography. The final prescription is reached by judging the dynamic lens movement, adequacy of the tear layer around the lens, corrected vision, and quality of vision. Various problems are encountered soon after prescribing lenses, including foreign body sensation, tear hypersecretion, decreased visual acuity, blurring, visual acuity change, redness, dryness, sudden pain, lens centering, and lens fallout. Here, we examine these problems and how to solve them.

Der Einfluss von blaulicht­reduzierenden Brillengläsern auf die Farbwahrnehmung

Purpose. To investigate the influence of two different blue- light-filtering lenses on subjective colour perception, using the Farnsworth-Munsell 100 Hue test. Material and Methods. Thirty subjects with normal colour vision performed the Farnsworth-Munsell 100 Hue test wearing three pairs of plano spectacle lenses. The total er- ror score was calculated with the “Farnsworth-Munsell Test Scoring Software“. The evaluated test lenses were the “Glacier Blue Shield“ (Shamir) and the “Wellness Protect 15“ (Eschen- bach) in addition to the 15 % grey tint (Hoya) as a reference lens. The transmission spectra of the lenses were measured beforehand with the spectrometer “Flame-S-UV-VIS-ES“ (“Ocean Optics“). Results. Using the univariate analysis of variance (ANOVA) with repeated measures, no significant difference could be detected in the colour discrimination capability between the reference lens from Hoya and the test lens from Sha- mir (p &gt; 0.05). A significant deterioration was found with the “Wellness Protect 15“ lens from Eschenbach compared to the reference lens from Hoya (p &lt; 0.001) and to the test lens from Shamir (p &lt; 0.001). In addition, a median disparity of the third box (43 – 63 of the Hue test) against all the other boxes was noticed for both the test lenses and the reference lens. Conclusion. With the blue-light-filtering anti-reflection coating from Shamir no difference in colour discrimination could be observed in this study. Only a stronger blue-light- filtering lens like “Wellness Protect 15“ leads to a significant deterioration. The closer the colours are in the CIE colour triangle, the bigger the confusion. To test the quality of colours with these lenses the colour rendering index should also be evaluated. Keywords Blue-light-filtering lenses, Colour discrimination, Colour perception, Colour Rendering Index, Transmission

The impact of spectacle lenses for myopia control on visual functions.

PurposeSpectacle lenses containing multiple small peripheral elements have been developed for myopia control in children. It is important that their effect on vision be quantified by (i) fixation through the peripheral portion, thereby using foveal vision and (ii) by fixation through the central portion and presentation of peripheral targets.MethodsThe above approaches were used in five studies to evaluate two novel spectacle lens designs: spectacle lenses with Highly Aspherical Lenslets (HAL) and Slightly Aspherical Lenslets (SAL). A single vision lens served as a control. Visually normal adults participated in each study. The first two studies had subjects fixate through the periphery of the lenses. High and low (10%) contrast visual acuity was measured with the Freiburg Vision Test and reading speed for high and low contrast words measured with a sentence generator. The other three studies assessed peripheral vision while subjects fixated through the central portion of the lens. Peripheral contrast sensitivity was measured using two cycles per degree drifting Gabor stimuli. Peripheral motion perception was further evaluated using random dot stimuli. Finally, attention was measured using an established test of useful field of view with three levels of complexity.ResultsThe periphery of the HAL lens significantly reduced low contrast visual acuity, but not high contrast visual acuity, while the effect of the SAL lens was not significant for either. Neither test lens affected reading speed for high contrast words, but the HAL lens significantly affected performance for low contrast words. Neither test lens affected peripheral motion perception or useful field of view.ConclusionsLow contrast visual acuity and reading was slightly reduced while high contrast visual acuity was unaffected when fixating through the periphery of the novel lens designs. None of the peripheral measures of vision was affected by the novel lens designs.

Large-scale Gravitational Lens Modeling with Bayesian Neural Networks for Accurate and Precise Inference of the Hubble Constant

Abstract We investigate the use of approximate Bayesian neural networks (BNNs) in modeling hundreds of time delay gravitational lenses for Hubble constant (H 0) determination. Our BNN was trained on synthetic Hubble Space Telescope quality images of strongly lensed active galactic nuclei with lens galaxy light included. The BNN can accurately characterize the posterior probability density functions (PDFs) of model parameters governing the elliptical power-law mass profile in an external shear field. We then propagate the BNN-inferred posterior PDFs into an ensemble H 0 inference, using simulated time delay measurements from a plausible dedicated monitoring campaign. Assuming well-measured time delays and a reasonable set of priors on the environment of the lens, we achieve a median precision of 9.3% per lens in the inferred H 0. A simple combination of a set of 200 test lenses results in a precision of 0.5 km s−1 Mpc−1 (0.7%), with no detectable bias in this H 0 recovery test. The computation time for the entire pipeline—including the generation of the training set, BNN training and H 0 inference—translates to 9 minutes per lens on average for 200 lenses and converges to 6 minutes per lens as the sample size is increased. Being fully automated and efficient, our pipeline is a promising tool for exploring ensemble-level systematics in lens modeling for H 0 inference.