-
Categories
-
Pharmaceutical Intermediates
-
Active Pharmaceutical Ingredients
-
Food Additives
- Industrial Coatings
- Agrochemicals
- Dyes and Pigments
- Surfactant
- Flavors and Fragrances
- Chemical Reagents
- Catalyst and Auxiliary
- Natural Products
- Inorganic Chemistry
-
Organic Chemistry
-
Biochemical Engineering
- Analytical Chemistry
- Cosmetic Ingredient
-
Pharmaceutical Intermediates
Promotion
ECHEMI Mall
Wholesale
Weekly Price
Exhibition
News
-
Trade Service
Report editor of the Heart of the Machine: A few years ago, Zenan and Xiaozhou, when a computer simultaneously recorded 10,000 neuron activities, it would become news.
In a blink of an eye, someone has already made one million.
In the science fiction novel "Three-Body", after eight years of hibernation by the Wallfacer Hines, mankind has developed a computer with computing power 10,000 times higher than that before hibernation, so that mankind can perform a panoramic analysis of the human brain.
Although there is no wall-facer plan yet, in the real world, the technology of scanning the brain has been developing, and its speed is faster than Moore's Law: in the past ten years, the number of simultaneous recordings of human cells has increased by nearly one.
Ten thousand times.
A recent study from the Vaziri Laboratory of Rockefeller University in the United States once again greatly increased the number of real-time computer detection cells to 1 million neurons.
At this rate, we may be able to resolve 100 million cells by 2030, and the number of neurons in the human brain is about 100 billion.
In the results of brain calcium imaging presented by Vaziri's laboratory, individual neurons flashed in sequence.
In the paper "High-Speed, Cortex-Wide Volumetric Recording of Neuroactivity at Cellular Resolution using Light Beads Microscopy" submitted to Biorxiv, the researchers realized the recording of scanning one million active neurons at a speed of 2 Hz, which is to allow The first step for the computer to learn what you are thinking is also an important step.
In a few decades, we may be able to achieve "thought mapping" and achieve psychological diagnosis at a near-real-time speed, making the treatment of diseases more targeted.
Two-photon microscopy and genetically coded calcium indicators have become standard tools for high-resolution imaging of neural activity in scattered brain tissue.
However, its various implementations have not overcome the inherent trade-off between speed and spatiotemporal sampling in principle.
This is indispensable for achieving medium-sized volumetric recording of neural activity at a speed compatible with cell resolution and resolving calcium instantaneously.
In this paper, the researchers proposed Light Beads Microscopy (LBM), which is a scalable and optimal acquisition method in time and space, limited only by fluorescence lifetime, in which a set of axially separated and time-different focal points The entire axial imaging range is recorded almost at the same time, so that volume measurement can be recorded at a speed of 1.
41 × 10^8 voxels per second.
Using LBM, the study demonstrated mesoscopic and volumetric imaging at multiple scales in the mouse cortex.
Including cell resolution recordings in a volume of approximately 3×5×0.
5 mm^3.
It contains> 200,000 neurons at about 5Hz, a population of about 1 million neurons in the range of about 5.
4×6×0.
5mm^3 at about 2Hz, and subcellular resolution at a higher speed (9.
6 Hz) Rate volume record.
The field of vision provided by LBM is unprecedented.
With the help of it, neural calculations based on the coding and information processing of the mammalian cerebral cortex can be discovered.
Paper link: https:// In the paper, the researcher demonstrated LBM: a high-speed optical acquisition technology for both mesoscopic and volume 2pM.
In LBM, the microscope scans a set of axially separated and temporally distinct focal points (ie, "beads"), rather than a single focal point (Figure 1a).
Beads record the information in the entire depth range (approximately 500 µm) of the sample during the dwell time of a single pixel, so the LBM captures the entire volume within the time required to scan a single plane.
In addition, by sampling optimized spatial sampling, LBM can expand the volume FOV to a mesoscopic scale while retaining a volume rate compatible with GCaMP.
The light beads here are formed by a cavity-based multiplexing method, which is called Many-fold Axial Multiplexing Module (MAxiMuM).
The uniqueness of MAxiMuM is that it can scale N to the limit caused by the fluorescence lifetime of GCaMP and the repetition frequency of the laser, and can control the relative power and position of each beam.
It provides 30 times the axial multiplexing, 141 MHz voxel acquisition rate, and 16 µm plane-to-plane axial spacing.
These conditions are for densely labeled tissue volumes with a fluorescence lifetime-limited information rate.
The sampling is optimized and compatible, using one pulse per voxel to maximize the SNR excitation, while using the entire inter-pulse time interval.
The researchers validated the LBM by performing in vivo imaging in the neocortex in mice that were awake and exhibited transgenic expression of GCaMP6s.
After using the optimized spatial sampling strategy, it can maintain a scan at a frame rate of about 5Hz in a volume of about 3×5×0.
5mm^3, solving the GCaMP transient problem.
The authors repositioned the FOV to include as many different regions as possible within a single cortical hemisphere, including SSp, PTLp, RSP, and VISp (Figure 2a).
In order to stimulate neural activity in more functional areas of FOV, researchers have also found some sensory stimulation methods such as vision.
The typical recording time range of this method is about 9-30 minutes, covering 1.
5-460,000 neurons. If scanning at a frequency of 2Hz within a volume of 5.
4×6×0.
5mm^3, the coverage of neurons can be extended to 1 million.
If you want to increase the scanning speed, it can reach up to 9.
6 Hz.
Million-level literature analysis, 100,000 words in-depth interpretation of the 2020-2021 global AI technology development trend report The report content covers the trend analysis of artificial intelligence top conferences, the conclusion of the overall technology trend development, and the six sub-fields (natural language processing, computer vision, robotics) Detailed explanations of technology development trend data and questionnaire conclusions related to automation technology, machine learning, intelligent infrastructure, data intelligence technology, and cutting-edge intelligence technology.
Finally, there are five-year breakthrough events in six major technical fields and complete data of Synced Indicator.
Identify the QR code below and purchase the report immediately.
© THE END For reprinting, please contact this official account for authorization to contribute or seek coverage: content@jiqizhixin.
com
In a blink of an eye, someone has already made one million.
In the science fiction novel "Three-Body", after eight years of hibernation by the Wallfacer Hines, mankind has developed a computer with computing power 10,000 times higher than that before hibernation, so that mankind can perform a panoramic analysis of the human brain.
Although there is no wall-facer plan yet, in the real world, the technology of scanning the brain has been developing, and its speed is faster than Moore's Law: in the past ten years, the number of simultaneous recordings of human cells has increased by nearly one.
Ten thousand times.
A recent study from the Vaziri Laboratory of Rockefeller University in the United States once again greatly increased the number of real-time computer detection cells to 1 million neurons.
At this rate, we may be able to resolve 100 million cells by 2030, and the number of neurons in the human brain is about 100 billion.
In the results of brain calcium imaging presented by Vaziri's laboratory, individual neurons flashed in sequence.
In the paper "High-Speed, Cortex-Wide Volumetric Recording of Neuroactivity at Cellular Resolution using Light Beads Microscopy" submitted to Biorxiv, the researchers realized the recording of scanning one million active neurons at a speed of 2 Hz, which is to allow The first step for the computer to learn what you are thinking is also an important step.
In a few decades, we may be able to achieve "thought mapping" and achieve psychological diagnosis at a near-real-time speed, making the treatment of diseases more targeted.
Two-photon microscopy and genetically coded calcium indicators have become standard tools for high-resolution imaging of neural activity in scattered brain tissue.
However, its various implementations have not overcome the inherent trade-off between speed and spatiotemporal sampling in principle.
This is indispensable for achieving medium-sized volumetric recording of neural activity at a speed compatible with cell resolution and resolving calcium instantaneously.
In this paper, the researchers proposed Light Beads Microscopy (LBM), which is a scalable and optimal acquisition method in time and space, limited only by fluorescence lifetime, in which a set of axially separated and time-different focal points The entire axial imaging range is recorded almost at the same time, so that volume measurement can be recorded at a speed of 1.
41 × 10^8 voxels per second.
Using LBM, the study demonstrated mesoscopic and volumetric imaging at multiple scales in the mouse cortex.
Including cell resolution recordings in a volume of approximately 3×5×0.
5 mm^3.
It contains> 200,000 neurons at about 5Hz, a population of about 1 million neurons in the range of about 5.
4×6×0.
5mm^3 at about 2Hz, and subcellular resolution at a higher speed (9.
6 Hz) Rate volume record.
The field of vision provided by LBM is unprecedented.
With the help of it, neural calculations based on the coding and information processing of the mammalian cerebral cortex can be discovered.
Paper link: https:// In the paper, the researcher demonstrated LBM: a high-speed optical acquisition technology for both mesoscopic and volume 2pM.
In LBM, the microscope scans a set of axially separated and temporally distinct focal points (ie, "beads"), rather than a single focal point (Figure 1a).
Beads record the information in the entire depth range (approximately 500 µm) of the sample during the dwell time of a single pixel, so the LBM captures the entire volume within the time required to scan a single plane.
In addition, by sampling optimized spatial sampling, LBM can expand the volume FOV to a mesoscopic scale while retaining a volume rate compatible with GCaMP.
The light beads here are formed by a cavity-based multiplexing method, which is called Many-fold Axial Multiplexing Module (MAxiMuM).
The uniqueness of MAxiMuM is that it can scale N to the limit caused by the fluorescence lifetime of GCaMP and the repetition frequency of the laser, and can control the relative power and position of each beam.
It provides 30 times the axial multiplexing, 141 MHz voxel acquisition rate, and 16 µm plane-to-plane axial spacing.
These conditions are for densely labeled tissue volumes with a fluorescence lifetime-limited information rate.
The sampling is optimized and compatible, using one pulse per voxel to maximize the SNR excitation, while using the entire inter-pulse time interval.
The researchers validated the LBM by performing in vivo imaging in the neocortex in mice that were awake and exhibited transgenic expression of GCaMP6s.
After using the optimized spatial sampling strategy, it can maintain a scan at a frame rate of about 5Hz in a volume of about 3×5×0.
5mm^3, solving the GCaMP transient problem.
The authors repositioned the FOV to include as many different regions as possible within a single cortical hemisphere, including SSp, PTLp, RSP, and VISp (Figure 2a).
In order to stimulate neural activity in more functional areas of FOV, researchers have also found some sensory stimulation methods such as vision.
The typical recording time range of this method is about 9-30 minutes, covering 1.
5-460,000 neurons. If scanning at a frequency of 2Hz within a volume of 5.
4×6×0.
5mm^3, the coverage of neurons can be extended to 1 million.
If you want to increase the scanning speed, it can reach up to 9.
6 Hz.
Million-level literature analysis, 100,000 words in-depth interpretation of the 2020-2021 global AI technology development trend report The report content covers the trend analysis of artificial intelligence top conferences, the conclusion of the overall technology trend development, and the six sub-fields (natural language processing, computer vision, robotics) Detailed explanations of technology development trend data and questionnaire conclusions related to automation technology, machine learning, intelligent infrastructure, data intelligence technology, and cutting-edge intelligence technology.
Finally, there are five-year breakthrough events in six major technical fields and complete data of Synced Indicator.
Identify the QR code below and purchase the report immediately.
© THE END For reprinting, please contact this official account for authorization to contribute or seek coverage: content@jiqizhixin.
com
