The sheer amount of information being processed arising from the high-density network of nerve cells in the brain, imposes severe restrictions with respect to the number of neurons that can be observed simultaneously. Many techniques focus on smaller nerve cell populations, and introduce wired electrode interfaces directly contacting neural tissue. These tethered methods however can lead to a host of detrimental issues, such as movement restriction affecting behaviour, and infections at the interface site. Due to this, devices capable of wirelessly transmitting large amounts of neural data are needed. To correlate neuron activity with behaviour, such devices are inserted into various animal models, preferably in vivo. The mouse model is particularly popular because of the familiarity of its genetics, physiology, and low upkeep costs. With the advent of such techniques as optogenetics, which allow increasingly precise light mediated activation of nerve cells, the preference for mice was further reinforced where they are the primary animal of choice. However, the majority of currently available wireless devices are too big and heavy to allow the freely moving behaviour of mice. Most of these use the RF based data transmission, and use larger energy storage units to supply the power required for higher data rates. The size limitations also restrict the maximum dimensions for an antenna that ca be used, further restricting the maximum available bandwidth. To propose a solution to these issues, the thesis outlines the design of low power devices that use visible light communication (VLC) to transmit neural data. As a pathway towards a fully wireless in vivo device we first develop an in vitro system, and interfaced with a 61-channel array containing rodent retinal tissue.The device was validated by sending neural data through the optical link, using an off the shelf LED. A fully wireless in vivo package was then developed. It was verified with test signals in a simulated environment, and a head fixed mouse at transmission distance of 20 cm. During in vivo recording, neural activity was stimulated using optogenetic techniques. The developed wireless system transmits 32 channels of uncompressed data (10.24 Mbps) in a 5g package. Strategies to transition to freely moving experiments and scale up channel count will be discussed. Finally, methods to readily reduce the weight will be outlined.
|Date of Award||19 Sep 2019|
- University Of Strathclyde
|Sponsors||EPSRC (Engineering and Physical Sciences Research Council)|
|Supervisor||Keith Mathieson (Supervisor) & Shuzo Sakata (Supervisor)|