Scientists have achieved a breakthrough in real-time analysis of large-scale neuronal imaging, enabling closed-loop brain research. By leveraging data processing techniques from astronomy and a hybrid FPGA-GPU architecture, they have successfully decoded neuronal activities and controlled external devices in real-time.
Real-Time Analysis of Neuronal Imaging Enables Closed-Loop Brain Research
( Credit to: Medicalxpress )
Scientists have achieved a breakthrough in real-time analysis of large-scale neuronal imaging, enabling closed-loop brain research. By leveraging data processing techniques from astronomy and a hybrid FPGA-GPU architecture, they have successfully decoded neuronal activities and controlled external devices in real-time.
( Credit to: Medicalxpress )
This advancement has opened up new possibilities for the development of optical brain-machine interfaces (BMI) and the understanding of neuronal activity characteristics suitable for BMI. It also paves the way for closed-loop whole-brain-scale research, utilizing techniques like virtual reality based on whole-brain cellular-resolution optical imaging and optogenetic control.
Overcoming Challenges in Whole-Brain Neuronal Imaging
Whole-brain neuronal activity imaging is a powerful tool for understanding the brain's principles, but the massive data processing requirements have hindered real-time analysis and closed-loop research. Inspired by rapid radio burst detection technology in astronomy, scientists developed an optical neural signal preprocessing system using the FX design.
They utilized FPGA programming to regularize signals from optical sensors and then sent them to a GPU-based real-time processing system for high-speed nonlinear registration, signal extraction, decoding, and obtaining feedback signals for controlling external devices.
Real-Time Optogenetic Stimulation for Enhanced Brain Activation
In a closed-loop experiment, scientists functionally clustered neurons in the whole brain and used the spontaneous activity of selected ensembles as a trigger signal for real-time optogenetic stimulation on target neuron ensembles. The closed-loop stimulation proved more effective in activating downstream brain areas compared to open-loop stimulation.
Modulating Brain States with Real-Time Visual Stimulation
By monitoring the activity of the locus coeruleus (LC) norepinephrinergic system in real-time, scientists applied visual stimulation during the excitatory phase of LC neurons representing the animal's awake state. This resulted in stronger responses of neurons across the brain, indicating the modulation of brain states on visual information processing.
Virtual Reality Driven by Neuronal Activities
Scientists achieved real-time dimensionality reduction of all brain neurons' activities, allowing the establishment of a virtual reality system directly driven by neuronal activities. The gain coupling between neuronal activity and the environment could be adjusted, enabling the neuron ensemble to adaptively adjust its output based on gain changes.
Implications and Future Developments
This breakthrough in real-time analysis of large-scale neuronal imaging opens up new possibilities for the development of optical brain-machine interfaces (BMI). By uncovering neuronal activity characteristics suitable for BMI and understanding their underlying mechanisms, more efficient optical BMI technologies can be developed.
Furthermore, this advancement paves the way for closed-loop whole-brain-scale research, utilizing techniques like virtual reality based on whole-brain cellular-resolution optical imaging and optogenetic control.