Publications by authors named "Gary C Lee"

Purpose: This study sought to identify the most effective testing program for detecting visual-field defects in mild-stage glaucoma with central visual-field defects.

Design: A multicenter, retrospective diagnostic testing evaluation.

Participants: The study involved 93 eyes (83 patients) with mild-stage glaucoma (median mean deviation [interquartile range]: -1.

View Article and Find Full Text PDF
Article Synopsis
  • The study introduces a student-teacher framework that trains lightweight deep learning models using unlabeled images to quickly detect important optical coherence tomography B-scans.
  • A total of 27 lightweight models were trained on about 70,000 expert-labeled B-scans, achieving validation accuracy between 89.6% and 95.1%, with the best models being significantly faster than a pre-trained ResNet50 teacher network.
  • By incorporating around 500,000 unlabeled B-scans in the training process, the models improved performance, with the best achieving similar sensitivity and specificity to the ResNet50, showcasing the potential of using readily available unlabeled data.
View Article and Find Full Text PDF

Purpose: To analyze the relationship between rates of false positive (FP) responses and standard automated perimetry results.

Design: Prospective multicenter cross-sectional study.

Methods: One hundred twenty-six patients with manifest or suspect glaucoma were tested with Swedish Interactive Thresholding Algorithm (SITA) Standard, SITA Fast, and SITA Faster at each of 2 visits.

View Article and Find Full Text PDF

Purpose: To describe a new time-saving threshold visual field-testing strategy-Swedish Interactive Thresholding Algorithm (SITA) Faster, which is intended to replace SITA Fast-and to report on a clinical evaluation of this new strategy.

Design: Description and validity analysis for modifications applied to SITA Fast.

Methods: Five centers tested 1 eye of each of 126 glaucoma and glaucoma suspect patients with SITA Faster, SITA Fast, and SITA Standard at each of 2 visits.

View Article and Find Full Text PDF

Purpose: To examine the performance of Guided Progression Analysis (GPA; Carl Zeiss Meditec, Dublin, CA) in spectral-domain optical coherence tomography (OCT) in detecting progressive thinning of ganglion cell-inner plexiform layer (GCIPL) and retinal nerve fiber layer (RNFL) in glaucoma.

Design: Longitudinal, observational study.

Participants: A total of 196 eyes of 123 primary open-angle glaucoma patients (mean follow-up, 5.

View Article and Find Full Text PDF

Optical coherence tomography (OCT) has become a standard tool in ophthalmology clinics for diagnosing many retinal diseases. Nonetheless, the technical and clinical communities still lack a standardized phantom that could aid in evaluating and normalizing the many protocols and systems used for diagnosis. Existing retinal phantoms are able to mimic the thickness and scattering properties of the retinal layers but are unable to model the morphology of the foveal pit, particularly the tapering of the retinal layers.

View Article and Find Full Text PDF

A solid immersion lens can be applied for high-resolution subsurface analysis of integrated circuits and other physical systems. We present a thorough analysis of the focal field distribution of a solid immersion lens system of arbitrary thickness. Cases of linearly and radially polarized illumination are examined and accurate expressions derived for the electric field in the image space.

View Article and Find Full Text PDF

Purpose: To evaluate a novel soft, lightweight cushion that can match the magnetic susceptibility of human tissue. The magnetic susceptibility difference between air and tissue produces field inhomogeneities in the B(0) field, which leads to susceptibility artifacts in magnetic resonance imaging (MRI) studies.

Materials And Methods: Pyrolytic graphite (PG) microparticles were uniformly embedded into a foam cushion to reduce or eliminate field inhomogeneities at accessible air and tissue interfaces.

View Article and Find Full Text PDF