Two-photon excitation fluorescence microscopy(TPM),owing to its capacity for subcellular resolution,intrinsic optical sectioning,and superior penetration depth in turbid samples,has revolutionized biomedical research....Two-photon excitation fluorescence microscopy(TPM),owing to its capacity for subcellular resolution,intrinsic optical sectioning,and superior penetration depth in turbid samples,has revolutionized biomedical research.However,its layer-by-layer scanning to form a three-dimensional image inherently limits the volumetric imaging speed and increases phototoxicity significantly.In this study,we develop a gradient excitation technique to accelerate TPM volumetric imaging.The axial positions of the fluorophores can be decoded from the intensity ratio of the paired images obtained by sequentially exciting the specimen with two axially elongated two-photon beams of complementary gradient intensities.We achieved a 0.63μm axial localization precision and demonstrate the flexibility of the gradient TPM on various sparsely labeled samples,including bead phantoms,mouse brain tissues,and live macrophages.Compared with traditional TPM,our technique improves volumetric imaging speed(by at least sixfold),decreases photobleaching(i.e.,less than 2.07±2.89%in 25 min),and minimizes photodamages.展开更多
Artificial neural networks have shown great proficiency in transforming low-resolution microscopic images into high-resolution images.However,training data remains a challenge,as large-scale open-source databases of m...Artificial neural networks have shown great proficiency in transforming low-resolution microscopic images into high-resolution images.However,training data remains a challenge,as large-scale open-source databases of microscopic images are rare,particularly 3D data.Moreover,the long training times and the need for expensive computational resources have become a burden to the research community.We introduced a deep-learning-based self-supervised volumetric imaging approach,which we termed“Self-Vision.”The self-supervised approach requires no training data,apart from the input image itself.The lightweight network takes just minutes to train and has demonstrated resolution-enhancing power on par with or better than that of a number of recent microscopybased models.Moreover,the high throughput power of the network enables large image inference with less postprocessing,facilitating a large field-of-view(2.45 mm×2.45 mm)using a home-built two-photon microscopy system.Self-Vision can recover images from fourfold undersampled inputs in the lateral and axial dimensions,dramatically reducing the acquisition time.Self-Vision facilitates the use of a deep neural network for 3D microscopy imaging,easing the demanding process of image acquisition and network training for current resolutionenhancing networks.展开更多
基金National Key Research and Development Program of China(2017YFC0110200)National Natural Science Foundation of China(81822023,81927803,91959121,92159104,82071972)+2 种基金Natural Science Foundation of Guangdong Province(2019A1515011746,2020B121201010)Scientific Instrument Innovation Team of Chinese Academy of Sciences(GJJSTD20180002)Shenzhen Basic Research Program(QCYJ20180507182432303,RCJC20200714114433058,ZDSY20130401165820357).
文摘Two-photon excitation fluorescence microscopy(TPM),owing to its capacity for subcellular resolution,intrinsic optical sectioning,and superior penetration depth in turbid samples,has revolutionized biomedical research.However,its layer-by-layer scanning to form a three-dimensional image inherently limits the volumetric imaging speed and increases phototoxicity significantly.In this study,we develop a gradient excitation technique to accelerate TPM volumetric imaging.The axial positions of the fluorophores can be decoded from the intensity ratio of the paired images obtained by sequentially exciting the specimen with two axially elongated two-photon beams of complementary gradient intensities.We achieved a 0.63μm axial localization precision and demonstrate the flexibility of the gradient TPM on various sparsely labeled samples,including bead phantoms,mouse brain tissues,and live macrophages.Compared with traditional TPM,our technique improves volumetric imaging speed(by at least sixfold),decreases photobleaching(i.e.,less than 2.07±2.89%in 25 min),and minimizes photodamages.
基金National Natural Science Foundation of China(62105353,81927803,82071972,91959121,92159104)Natural Science Foundation of Guangdong Province(2019A1515011746,2020B121201010,2021A1515012022)+1 种基金Scientific Instrument Innovation Team of Chinese Academy of Sciences(GJJSTD20180002)Shenzhen Basic Research Program(RCJC20200714114433058,RCYX20210609104445093,ZDSY20130401165820357)。
文摘Artificial neural networks have shown great proficiency in transforming low-resolution microscopic images into high-resolution images.However,training data remains a challenge,as large-scale open-source databases of microscopic images are rare,particularly 3D data.Moreover,the long training times and the need for expensive computational resources have become a burden to the research community.We introduced a deep-learning-based self-supervised volumetric imaging approach,which we termed“Self-Vision.”The self-supervised approach requires no training data,apart from the input image itself.The lightweight network takes just minutes to train and has demonstrated resolution-enhancing power on par with or better than that of a number of recent microscopybased models.Moreover,the high throughput power of the network enables large image inference with less postprocessing,facilitating a large field-of-view(2.45 mm×2.45 mm)using a home-built two-photon microscopy system.Self-Vision can recover images from fourfold undersampled inputs in the lateral and axial dimensions,dramatically reducing the acquisition time.Self-Vision facilitates the use of a deep neural network for 3D microscopy imaging,easing the demanding process of image acquisition and network training for current resolutionenhancing networks.