Traditional imaging cytometry uses fluorescence markers to identify specific structures but is limited in throughput by the labeling process. We develop a label-free technique that alleviates the physical staining and provides multiplexed readouts via a deep learning–augmented digital labeling method. We leverage the rich structural information and superior sensitivity in reflectance microscopy and show that digital labeling predicts accurate subcellular features after training on immunofluorescence images. We demonstrate up to three times improvement in the prediction accuracy over the state of the art. Beyond fluorescence prediction, we demonstrate that single cell–level structural phenotypes of cell cycles are correctly reproduced by the digital multiplexed images, including Golgi twins, Golgi haze during mitosis, and DNA synthesis. We further show that the multiplexed readouts enable accurate multiparametric single-cell profiling across a large cell population. Our method can markedly improve the throughput for imaging cytometry toward applications for phenotyping, pathology, and high-content screening.
Purpose To evaluate the clinical utility of visible light optical coherence tomography (VIS-OCT) and to test whether VIS-OCT reflectivity and spectroscopy of peripapillary retinal nerve fiber layer (pRNFL) are correlated with severity of glaucoma, compared with standard-of-care OCT thickness measurements. Methods In total 54 eyes (20 normal, 17 suspect/preperimetric glaucoma [GS/PPG], 17 perimetric glaucoma [PG]) were successfully imaged with complete datasets. All the eyes were scanned by a custom-designed dual-channel device that simultaneously acquired VIS-OCT and near-infrared OCT (NIR-OCT) images. A 5 × 5 mm 2 scan was taken of the pRNFL. The pRNFL reflectivity was calculated for both channels and the spectroscopic marker was quantified by pVN, defined as the ratio of VIS-OCT to NIR-OCT pRNFL reflectivity. The results were compared with ophthalmic examinations and Zeiss Cirrus OCT. Results VIS-OCT pRNFL reflectivity significantly, sequentially decreased from normal to GS/PPG to PG, as did NIR-OCT pRNFL reflectivity. The pVN had the same decreasing trend among three groups. Normal and GS/PPG eyes were significantly different in VIS-OCT pRNFL reflectivity ( P = 0.002) and pVN ( P < 0.001), but were not in NIR-OCT pRNFL reflectivity ( P = 0.14), circumpapillary RNFL thickness ( P = 0.17), or macular ganglion cell layer and inner plexiform layer thickness ( P = 0.07) in a mixed linear regression model. Conclusions VIS-OCT pRNFL reflectivity and pVN better distinguished GS/PPG from normal eyes than Cirrus OCT thickness measurements. Translational Relevance VIS-OCT pRNFL reflectivity and pVN could be useful metrics in the early detection of glaucoma upon further longitudinal validation.
Traditional imaging cytometry uses fluorescence markers to identify specific structures, but is limited in throughput by the labeling process. Here we develop a label-free technique that alleviates the physical staining and provides highly multiplexed readout via a deep learning-augmented digital labeling method. We leverage the rich structural information and superior sensitivity in reflectance microscopy and show that digital labeling predicts highly accurate subcellular features after training on immunofluorescence images. We demonstrate up to 3× improvement in the prediction accuracy over the state-of-the-art. Beyond fluorescence prediction, we demonstrate that single-cell level structural phenotypes of cell cycles are correctly reproduced by the digital multiplexed images, including Golgi twins, Golgi haze during mitosis and DNA synthesis. We further show that the multiplexed readout enables accurate multi-parametric single-cell profiling across a large cell population. Our method can dramatically improve the throughput for imaging cytometry towards applications for phenotyping, pathology, and high-content screening.
Purpose: To evaluate the clinical utility of visible light optical coherence tomography (VIS-OCT) and to test whether VIS-OCT reflectivity and spectroscopy of peripapillary retinal nerve fiber layer (pRNFL) are correlated with severity of glaucoma, compared with standard-of-care OCT thickness measurements. Design: Cross-sectional study. Method: Fifty-four eyes from three groups of subjects (normal, glaucoma suspect, and glaucoma) were scanned by a custom-designed dual-channel device that simultaneously acquired visible (VIS-OCT) and near-infrared OCT (NIR-OCT) images. A 5x5 mm2 scan was taken of the peripapillary nerve fiber layer (pRNFL). The pRNFL reflectivity was calculated for both channels and the spectroscopic marker was quantified by pVN, defined by the ratio of VIS-OCT to NIR-OCT pRNFL reflectivity. The results were compared with ophthalmic exams and clinical Zeiss OCT measurements. Mixed linear model was used to evaluate the association of imaging markers with glaucoma severity. Results: VIS-OCT pRNFL reflectivity significantly, sequentially declined from normal to suspect to glaucoma (3.42 ± 0.35, 2.58 ± 0.37, 2.16 ± 0.39, mean ± SD), as did NIR-OCT pRNFL reflectivity (2.46 ± 0.24, 2.27 ± 0.24, 1.96 ± 0.25). The pVN also had the same decreasing trend among three groups (1.39 ± 0.08, 1.14 ± 0.08, 1.08 ± 0.10). Normal and suspect eyes were significantly different in VIS-OCT pRNFL reflectivity (p=0.002), pVN (p<0.001) but not in NIR-OCT pRNFL reflectivity (p=0.14), circumpapillary RNFL (cpRNFL) thickness (p=0.17), or macular ganglion cell layer and inner plexiform layer (GCL+IPL) thickness (p=0.07) in the mixed linear model. Conclusion: VIS-OCT pRNFL reflectivity and pVN was more sensitive in separating suspect eyes from normal ones than clinical OCT thickness measurements. VIS-OCT pRNFL reflectivity and pVN could be a useful metric in early detection of glaucoma upon further longitudinal validation.
A computational alternative to standard immunofluorescence (IF) imaging based on deep learning model is proposed for transforming morphological information from reflectance microscopy to specific and accurate IF predictions with high multiplicity.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.