A deep learning network for parallel self-denoising and segmentation in visible light optical coherence tomography of human retina DOI Creative Commons
Tianyi Ye, Jingyu Wang, Ji Yi

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2022, Volume and Issue: unknown

Published: Nov. 27, 2022

Abstract Visible light optical coherence tomography (VIS-OCT) of human retina is an emerging imaging modality that uses shorter wavelength in visible range than conventional near infrared (NIR) light. It provides one-micron level axial resolution to better separate stratified retinal layers, as well microvascular oximetry. However, due the practical limitation laser safety and comfort, permissible illumination power much lower NIR OCT which can be challenging obtain high quality VIS-OCT images subsequent image analysis. Therefore, improving by denoising essential step overall workflow clinical applications. In this paper, we provide first dataset from normal eyes, including layer annotation “noisy-clean” pairs. We propose efficient co-learning deep learning framework for parallel self-denoising segmentation simultaneously. Both tasks synergize within same network improve each other’s performance. The significant improvement (2% higher Dice coefficient compared segmentation-only process) ganglion cell (GCL), inner plexiform (IPL) nuclear (INL) observed when available drops 25%, suggesting annotation-efficient training. also showed model trained on our generalizes a different scanning protocol.

Language: Английский

Photoreceptor Outer Segment Reflectivity With Ultrahigh-Resolution Visible-Light Optical Coherence Tomography in Systemic Hydroxychloroquine Use DOI Creative Commons
Anupam Garg, Jingyu Wang,

Bailee Alonzo

et al.

Translational Vision Science & Technology, Journal Year: 2025, Volume and Issue: 14(3), P. 2 - 2

Published: March 3, 2025

To evaluate outer retinal organization in normal subjects and those using hydroxychloroquine (HCQ) with ultrahigh-resolution visible-light optical coherence tomography (VIS-OCT). Forty eyes of 22 adult were recruited from a tertiary-care retina practice, including controls (20 eyes, 12 subjects, 40 ± years old) history HCQ use 10 62 17 old). Each subject was imaged custom-built VIS-OCT device (axial resolution 1.3 µm) U.S. Food Drug Administration-approved OCT devices. With the VIS-OCT, control demonstrated five six hyperreflective bands foveal parafoveal regions, respectively, between nuclear layer Bruch's membrane. These intensity profiles complementary to known histopathologic distribution rods cones. In comparison controls, taking reduced all bands, particularly two four. cases suspected or toxicity, attenuation band 3i, no there other that more severe than suggesting changes reflectivity 3i may be earliest identifiable sign toxicity. revealed unique banding pattern corresponding photoreceptor density distributions. Notable segment profile associated early This finding an early, possibly reversible, primarily impacting is useful detecting subclinical structural found hydroxychloroquine.

Language: Английский

Citations

1
可见光光学相干层析成像技术发展综述（特邀） DOI

宋维业 Song Weiye,

姚政开 Yao Zhengkai,

吴付旺 Wu Fuwang

et al.

Chinese Journal of Lasers, Journal Year: 2024, Volume and Issue: 51(15), P. 1507101 - 1507101

Published: Jan. 1, 2024

Citations

2
Band Visibility in High-Resolution Optical Coherence Tomography Assessed With a Custom Review Tool and Updated, Histology-Derived Nomenclature DOI Creative Commons
Lukas Goerdt, Thomas A. Swain, Deepayan Kar

et al.

Translational Vision Science & Technology, Journal Year: 2024, Volume and Issue: 13(12), P. 19 - 19

Published: Dec. 13, 2024

Purpose: For structure-function research at the transition of aging to age-related macular degeneration, we refined current consensus optical coherence tomography (OCT) nomenclature and evaluated a novel review software for investigational high-resolution OCT imaging (HR-OCT; <3 µm axial resolution). Method: Volume electron microscopy, immunolocalizations, histology, devices informed custom ImageJ-based tool assess retinal band visibility. We examined effects on visibility automated real-time averaging (ART) 9 100 (11 eyes 10 healthy young adults), (10 vs 22 aged), degeneration (AMD; aged, 17 early (e)AMD, 15 intermediate (i)AMD). Intrareader reliability was assessed. Results: Bands not included in are now visible using HR-OCT: inner plexiform layer (IPL) 1-5, outer (OPL) 1-2, segment interdigitation zone 1-2 (OSIZ, including hyporeflective segments), pigment epithelium (RPE) 1-5. Cohen's kappa 0.54–0.88 0.67–0.83 bands subset eyes. IPL-3-5 OPL-2 benefitted from increased ART. OSIZ-2 RPE-1,2,3,5 worse aged than OSIZ-1-2, RPE-1, RPE-5 decreased eAMD iAMD compared Conclusions: reliably identified 28 HR-OCT. Image improved Aging AMD development impacted Translational Significance: Detailed knowledge anatomic structures will enhance precision research, AI training analyses.

Language: Английский

Citations

2
Deep learning network for parallel self-denoising and segmentation in visible light optical coherence tomography of the human retina DOI Creative Commons
Tianyi Ye, Jingyu Wang, Ji Yi

et al.

Biomedical Optics Express, Journal Year: 2023, Volume and Issue: 14(11), P. 6088 - 6088

Published: Oct. 20, 2023

Visible light optical coherence tomography (VIS-OCT) of the human retina is an emerging imaging modality that uses shorter wavelengths in visible range than conventional near-infrared (NIR) light. It provides one-micron level axial resolution to better separate stratified retinal layers, as well microvascular oximetry. However, due practical limitation laser safety and comfort, permissible illumination power much lower NIR OCT, which can be challenging obtain high-quality VIS-OCT images subsequent image analysis. Therefore, improving quality by denoising essential step overall workflow clinical applications. In this paper, we provide first dataset from normal eyes, including layer annotation “noisy-clean” pairs. We propose efficient co-learning deep learning framework for parallel self-denoising segmentation simultaneously. Both tasks synergize within same network improve each other’s performance. The significant improvement (2% higher Dice coefficient compared segmentation-only process) ganglion cell (GCL), inner plexiform (IPL) nuclear (INL) observed when available drops 25%, suggesting annotation-efficient training. also showed model trained on our generalizes a different scanning protocol.

Language: Английский

Citations

4
BreakNet: Discontinuity-Resilient Multi-Scale Transformer Segmentation of Retinal Layers DOI Creative Commons
Razieh Ganjee, Bingjie Wang, Lingyun Wang

et al.

Biomedical Optics Express, Journal Year: 2024, Volume and Issue: 15(12), P. 6725 - 6725

Published: Oct. 30, 2024

Visible light optical coherence tomography (vis-OCT) is gaining traction for retinal imaging due to its high resolution and functional capabilities. However, the significant absorption of hemoglobin in visible range leads pronounced shadow artifacts from blood vessels, posing challenges accurate layer segmentation. In this study, we present BreakNet, a multi-scale Transformer-based segmentation model designed address boundary discontinuities caused by these artifacts. BreakNet utilizes hierarchical Transformer convolutional blocks extract global local feature maps, capturing essential contextual, textural, edge characteristics. The incorporates decoder that expand pathways enhance extraction fine details semantic information, ensuring precise Evaluated on rodent images acquired with prototype vis-OCT, demonstrated superior performance over state-of-the-art models, such as TCCT-BP U-Net, even when faced limited-quality ground truth data. Our findings indicate has potential significantly improve quantification analysis.

Language: Английский

Citations

1
The effects of time restricted feeding on age-related changes in the mouse retina DOI Creative Commons

Cade A. Huston,

Madison Milan,

Michaela L. Vance

et al.

Experimental Gerontology, Journal Year: 2024, Volume and Issue: 194, P. 112510 - 112510

Published: July 5, 2024

Dietary modifications such as caloric restriction (CR) and intermittent fasting (IF) have gained popularity due to their proven health benefits in aged populations. In time restricted feeding (TRF), a form of fasting, the amount for food intake is regulated without restricting intake. TRF beneficial central nervous system support brain context aging. Therefore, we here ask whether also exerts effects retina. We compared mice (24 months) on paradigm (access six hours per day) either 6 or 12 months against young control (8 an ad libitum diet. examined changes retina at functional (electroretinography), structural (histology fluorescein angiograms) molecular (gene expression) level. treatment showed amelioration age-related reductions both scotopic photopic b-wave amplitudes suggesting retinal interneuron signaling. did not affect signs inflammation microglial activation histological Our data indicate that helps preserve some aspects function are decreased with aging, adding our understanding altered patterns may confer.

Language: Английский

Citations

0
Enhanced Framework for Concurrent correction and Segmentation in Retinal Optical Coherence Tomography DOI

M. Pavithra,

D Archana,

A Tamilkumaran

et al.

2022 International Conference on Communication, Computing and Internet of Things (IC3IoT), Journal Year: 2024, Volume and Issue: unknown, P. 1 - 6

Published: April 17, 2024

A revolutionary imaging technique called optical coherence tomography of visible light (VIS-OCT) the individual uses shorter wavelengths than traditional near-infrared (NIR) light. To more accurately discern stratified retinal layers, it offers microvascular oximetry together with one-micron level axial resolution. Since allowed illumination power is significantly lower NIR OCT due to practical limits regarding laser safety and comfort, might be difficult generate VIS-OCT images a high enough quality for further image processing. As result, denoising crucial step in whole workflow clinical applications involving VIS-OCT. The first collection presented this study from normal eyes obtained using We offer simultaneous self-denoising segmentation system based on deep learning. Both tasks complement one another inside same network increase each other's productivity. Annotation-efficient training demonstrated by discernible When available annotation falls below 25%, Dice coefficient was 2% higher segmentation-only method.) accomplished. Additionally, we how well model learned our dataset could applied an alternative scanning method.

Language: Английский

Citations

0
Photoreceptor outer segment reflectivity with ultrahigh resolution visible light optical coherence tomography in systemic hydroxychloroquine use DOI Creative Commons
Anupam Garg, Jingyu Wang,

Bailee Alonzo

et al.

medRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: Sept. 12, 2024

To evaluate outer retinal organization in normal subjects and those using hydroxychloroquine (HCQ) with ultrahigh resolution visible light optical coherence tomography (VIS-OCT).

Language: Английский

Citations

0
A deep learning network for parallel self-denoising and segmentation in visible light optical coherence tomography of human retina DOI Creative Commons
Tianyi Ye, Jingyu Wang, Ji Yi

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2022, Volume and Issue: unknown

Published: Nov. 27, 2022

Abstract Visible light optical coherence tomography (VIS-OCT) of human retina is an emerging imaging modality that uses shorter wavelength in visible range than conventional near infrared (NIR) light. It provides one-micron level axial resolution to better separate stratified retinal layers, as well microvascular oximetry. However, due the practical limitation laser safety and comfort, permissible illumination power much lower NIR OCT which can be challenging obtain high quality VIS-OCT images subsequent image analysis. Therefore, improving by denoising essential step overall workflow clinical applications. In this paper, we provide first dataset from normal eyes, including layer annotation “noisy-clean” pairs. We propose efficient co-learning deep learning framework for parallel self-denoising segmentation simultaneously. Both tasks synergize within same network improve each other’s performance. The significant improvement (2% higher Dice coefficient compared segmentation-only process) ganglion cell (GCL), inner plexiform (IPL) nuclear (INL) observed when available drops 25%, suggesting annotation-efficient training. also showed model trained on our generalizes a different scanning protocol.

Language: Английский

Citations

1