Dataset Information

High-Throughput Molecular Imaging via Deep-Learning-Enabled Raman Spectroscopy.

ABSTRACT: Raman spectroscopy enables nondestructive, label-free imaging with unprecedented molecular contrast, but is limited by slow data acquisition, largely preventing high-throughput imaging applications. Here, we present a comprehensive framework for higher-throughput molecular imaging via deep-learning-enabled Raman spectroscopy, termed DeepeR, trained on a large data set of hyperspectral Raman images, with over 1.5 million spectra (400 h of acquisition) in total. We first perform denoising and reconstruction of low signal-to-noise ratio Raman molecular signatures via deep learning, with a 10× improvement in the mean-squared error over common Raman filtering methods. Next, we develop a neural network for robust 2-4× spatial super-resolution of hyperspectral Raman images that preserve molecular cellular information. Combining these approaches, we achieve Raman imaging speed-ups of up to 40-90×, enabling good-quality cellular imaging with a high-resolution, high signal-to-noise ratio in under 1 min. We further demonstrate Raman imaging speed-up of 160×, useful for lower resolution imaging applications such as the rapid screening of large areas or for spectral pathology. Finally, transfer learning is applied to extend DeepeR from cell to tissue-scale imaging. DeepeR provides a foundation that will enable a host of higher-throughput Raman spectroscopy and molecular imaging applications across biomedicine.

SUBMITTER: Horgan CC

PROVIDER: S-EPMC9286315 | biostudies-literature | 2021 Dec

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

High-Throughput Molecular Imaging via Deep-Learning-Enabled Raman Spectroscopy.

Horgan Conor C CC Jensen Magnus M Nagelkerke Anika A St-Pierre Jean-Philippe JP Vercauteren Tom T Stevens Molly M MM Bergholt Mads S MS

Analytical chemistry 20211119 48

Raman spectroscopy enables nondestructive, label-free imaging with unprecedented molecular contrast, but is limited by slow data acquisition, largely preventing high-throughput imaging applications. Here, we present a comprehensive framework for higher-throughput molecular imaging via deep-learning-enabled Raman spectroscopy, termed DeepeR, trained on a large data set of hyperspectral Raman images, with over 1.5 million spectra (400 h of acquisition) in total. We first perform denoising and reco ...[more]

PMID: 34797972

Similar Datasets

Project description:Optical fiber probe-based Raman spectroscopy systems are widely used for in situ measurements ranging from material characterization to biomedical applications. However, small Raman cross sections necessitate the use of high-power lasers or long exposure times that limit Raman's larger application to multiple research fields. This limitation can be overcome by collecting more Raman photons through additional collection fibers with taller detectors. This system configuration requires replacement of the detector and modification of the spectrograph to incorporate larger optical components, making it a costly and cumbersome option. In probe-based Raman systems, a typical detector image shows stacked collection fibers on the vertical axis and Raman spectra on the horizontal axis. While the vertical pixels are fully packed with multiple collection fibers, horizontal pixels have broad silent regions due to the narrow bandwidth of Raman peaks, potentially wasting valuable detector pixels. Here, we propose a new approach utilizing horizontally shifted collection fibers rather than vertically stacked ones. We designed and fabricated a novel collection fiber bundle that has horizontally shifted optical fibers in two vertical lines at the spectrograph entrance. This custom-made fiber bundle was incorporated into the imaging spectrograph to provide multiple horizontally shifted spectra on the detector. Through deconvolution, the original spectra can be recovered with an improved detection limit from greater photon collection. We demonstrate an enhanced limit of detection on various bioanalytes, such as glucose, urea, and lactate. Further, we applied the probe to measure tissue Raman spectra and successfully decomposed them into basis spectra, demonstrating the potential application of high-throughput in vivo tissue diagnosis. Our approach provides a simple, cost-effective, and universal method to increase the throughput without modifying existing Raman spectrometers.

Dataset Information

High-Throughput Molecular Imaging via Deep-Learning-Enabled Raman Spectroscopy.

Publications

High-Throughput Molecular Imaging via Deep-Learning-Enabled Raman Spectroscopy.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets