Brain Machine Interface

One of the main project that I have created for myself is a brain machine interface.

Lots of it is theory and design.

I have started work on a USB PCR machine. The reason for working on a PCR machine is that I can create a system that uses analog inputs and digital outputs. The use of analog inputs and digital outputs will help learn to interface with neurons.

It will allow me to learn about most about the components I will need to create a Brain Machine Interface(BMI).

I have used an analog to digital converter, and converted temperature reading to digital signals. Configured a USBee ex test pod to take the digitized inputs and transfer them to the computer. The computer then sends a signal to a digital to analog converter that controls a Peltier device for altering the temperature in a water.

Having such a system will allow me to alter the temperature of a water bath for DNA replication. The PRC controls can also be adapted for environmental controls for nematodes, diatoms and whatever else may need a temperature controlled chamber in the future.

With PCR machine working and chambers to control the growth of E.Coli and nematodes, work on opto-genetics can start and the ion channels can be recombine to form fusion proteins that will allow light to activate the channels and alter the membrane potential of the cells they are expressed in.

Also a more difficult task of getting a light emitting protein like green fluorescent protein(GFP) to be linked to a voltage sensitive protein that can accurately sense and emit light based on the cell’s membrane potential.

The genes for the light emitting proteins and the channelrhodopsin can be introduced with a benign vector such as a lentivirus engineered to carry them in to the brain.

Ideally if a visual processor called the octopus retina would read the light emitted from the neurons. The light information is transformed in to spike frequencies by the octopus retina camera. This spike encoding is very similar to that of the neurons. because it is similar the neurons should be able to interpret it and use that information for their own processing.

The spike data could then be passes on to a neuromorphic chip optimized to run a neural net. The neural net running on the chip would then analyse the input and create an appropriate output that could be feed back in to the brain or passed on to a different computer to do alternative processing.

To feed the information back into the brain. The use of optical fiber has been employed to trigger different bacterial channelrhodopsin

that have been transferred in to many different animal species. There are many wavelengths of light that can trigger a channelrhodopsin . The channelrhodopsin have also been altered to function with different time frames and even with g coupled proteins. Because of the variety of channelrhodopsin there is many different ways that the neurons can be affected and even different populations of neurons affected in different ways.

Once the channelrhodopsin has been triggered with pulse of light the neurons alter their firing pattern and this information has been fed in to the brain.

I could go into more detail here for any neuroscientists out there (perhaps in a future post). Send me a message if there is anything that you would like added or you think should be changed. Thank you.

This was an old post that I figured I should carry forward.

I have much more to say on this and new ideas I'll share with you soon.