CellSighter: a neural network to classify cells in highly multiplexed images

Nature Communications, Journal Year: 2023, Volume and Issue: 14(1)

Published: July 18, 2023

Abstract Multiplexed imaging enables measurement of multiple proteins in situ, offering an unprecedented opportunity to chart various cell types and states tissues. However, classification, the task identifying type individual cells, remains challenging, labor-intensive, limiting throughput. Here, we present CellSighter, a deep-learning based pipeline accelerate classification multiplexed images. Given small training set expert-labeled images, CellSighter outputs label probabilities for all cells new achieves over 80% accuracy major across platforms, which approaches inter-observer concordance. Ablation studies simulations show that is able generalize its data learn features protein expression levels, as well spatial such subcellular patterns. CellSighter’s design reduces overfitting, it can be trained with only thousands or even hundreds labeled examples. also prediction confidence, allowing downstream experts control results. Altogether, drastically hands-on time while improving consistency datasets.

Language: Английский

---

Two-photon synthetic aperture microscopy for minimally invasive fast 3D imaging of native subcellular behaviors in deep tissue

Cell, Journal Year: 2023, Volume and Issue: 186(11), P. 2475 - 2491.e22

Published: May 1, 2023

Holistic understanding of physio-pathological processes requires noninvasive 3D imaging in deep tissue across multiple spatial and temporal scales to link diverse transient subcellular behaviors with long-term physiogenesis. Despite broad applications two-photon microscopy (TPM), there remains an inevitable tradeoff among spatiotemporal resolution, volumes, durations due the point-scanning scheme, accumulated phototoxicity, optical aberrations. Here, we harnessed concept synthetic aperture radar TPM achieve aberration-corrected dynamics at a millisecond scale for over 100,000 large volumes tissue, three orders magnitude reduction photobleaching. With its advantages, identified direct intercellular communications through migrasome generation following traumatic brain injury, visualized formation process germinal center mouse lymph node, characterized heterogeneous cellular states visual cortex, opening up horizon intravital understand organizations functions biological systems holistic level.

Language: Английский

---

Zero-shot learning enables instant denoising and super-resolution in optical fluorescence microscopy

Nature Communications, Journal Year: 2024, Volume and Issue: 15(1)

Published: May 16, 2024

Abstract Computational super-resolution methods, including conventional analytical algorithms and deep learning models, have substantially improved optical microscopy. Among them, supervised neural networks demonstrated outstanding performance, however, demanding abundant high-quality training data, which are laborious even impractical to acquire due the high dynamics of living cells. Here, we develop zero-shot deconvolution (ZS-DeconvNet) that instantly enhance resolution microscope images by more than 1.5-fold over diffraction limit with 10-fold lower fluorescence ordinary imaging conditions, in an unsupervised manner without need for either ground truths or additional data acquisition. We demonstrate versatile applicability ZS-DeconvNet on multiple modalities, total internal reflection microscopy, three-dimensional wide-field confocal two-photon lattice light-sheet multimodal structured illumination enables multi-color, long-term, 2D/3D subcellular bioprocesses from mitotic single cells multicellular embryos mouse C. elegans .

Language: Английский

---

Multiphoton fluorescence microscopy for in vivo imaging

Cell, Journal Year: 2024, Volume and Issue: 187(17), P. 4458 - 4487

Published: Aug. 1, 2024

Language: Английский

---

Bridging live-cell imaging and next-generation cancer treatment

Nature reviews. Cancer, Journal Year: 2023, Volume and Issue: 23(11), P. 731 - 745

Published: Sept. 13, 2023

Language: Английский

---

Statistically unbiased prediction enables accurate denoising of voltage imaging data

Nature Methods, Journal Year: 2023, Volume and Issue: 20(10), P. 1581 - 1592

Published: Sept. 18, 2023

Abstract Here we report SUPPORT (statistically unbiased prediction utilizing spatiotemporal information in imaging data), a self-supervised learning method for removing Poisson–Gaussian noise voltage data. is based on the insight that pixel value data highly dependent its neighboring pixels, even when temporally adjacent frames alone do not provide useful statistical prediction. Such dependency captured and used by convolutional neural network with blind spot to accurately denoise which existence of action potential time frame cannot be inferred other frames. Through simulations experiments, show enables precise denoising types microscopy image while preserving underlying dynamics within scene.

Language: Английский

---

Recent advances in carbon dots-based nanoplatforms: Physicochemical properties and biomedical applications

Chemical Engineering Journal, Journal Year: 2023, Volume and Issue: 476, P. 146593 - 146593

Published: Oct. 10, 2023

Language: Английский

---

Spatial redundancy transformer for self-supervised fluorescence image denoising

Nature Computational Science, Journal Year: 2023, Volume and Issue: 3(12), P. 1067 - 1080

Published: Dec. 11, 2023

Abstract Fluorescence imaging with high signal-to-noise ratios has become the foundation of accurate visualization and analysis biological phenomena. However, inevitable noise poses a formidable challenge to sensitivity. Here we provide spatial redundancy denoising transformer (SRDTrans) remove from fluorescence images in self-supervised manner. First, sampling strategy based on is proposed extract adjacent orthogonal training pairs, which eliminates dependence speed. Second, designed lightweight spatiotemporal architecture capture long-range dependencies high-resolution features at low computational cost. SRDTrans can restore high-frequency information without producing oversmoothed structures distorted traces. Finally, demonstrate state-of-the-art performance single-molecule localization microscopy two-photon volumetric calcium imaging. does not contain any assumptions about process sample, thus be easily extended various modalities applications.

Language: Английский

---

Self-supervised denoising for multimodal structured illumination microscopy enables long-term super-resolution live-cell imaging

PhotoniX, Journal Year: 2024, Volume and Issue: 5(1)

Published: March 1, 2024

Abstract Detection noise significantly degrades the quality of structured illumination microscopy (SIM) images, especially under low-light conditions. Although supervised learning based denoising methods have shown prominent advances in eliminating noise-induced artifacts, requirement a large amount high-quality training data severely limits their applications. Here we developed pixel-realignment-based self-supervised framework for SIM (PRS-SIM) that trains an image denoiser with only noisy and substantially removes reconstruction artifacts. We demonstrated PRS-SIM generates artifact-free images 20-fold less fluorescence than ordinary imaging conditions while achieving comparable super-resolution capability to ground truth (GT). Moreover, easy-to-use plugin enables both implementation multimodal platforms including 2D/3D linear/nonlinear SIM. With PRS-SIM, achieved long-term live-cell various vulnerable bioprocesses, revealing clustered distribution Clathrin-coated pits detailed interaction dynamics multiple organelles cytoskeleton.

Language: Английский

---

Readily releasable β cells with tight Ca2+–exocytosis coupling dictate biphasic glucose-stimulated insulin secretion

Nature Metabolism, Journal Year: 2024, Volume and Issue: 6(2), P. 238 - 253

Published: Jan. 26, 2024

Language: Английский

---