“If you take my equipment away from me, I might as well retire”, John O’Keefe, Nobel Prize in Physiology or Medicine, 2014.
I General introduction
The open-source community popularizes knowledge in a scale never seen before. In regards to scientific equipment this popularization first reduces prices, but moreover, allows the researcher to better look, control and understand the different elements in the recording set-up providing the signals observed from the experiment. Such an understanding is a prerequisite in formulating any scientific conclusion, but also provides a perfect case for education and teaching purposes.
The part of the open-source community, whom is interested in electrophysiology, has already produced three different equipment for neuro-scientific related measurements: “Backyard Brain” for insect electrophysiology, “OpenBCI” for EEG on humans, and “Open-Ephys” for in vivo neuroscience. Surprisingly, nothing exists now for neuronal cells recording in culture (or insect explants). This project proposes to build an open-source amplifier to monitor neuronal population in culture, i.e. neuronal cells grown together in a Petri dish. As reported in the literature, building a complete set-up, including an amplifier, may cost approximately 30k$ (Chiang Liao … Chen 2012), while Backyard Brain system cost 400$, OpenBCI system 1.5k$, and Open-Ephys 5k$.
Reducing the cost of an amplifier for neurons in culture will help in popularizing neuroscience. Hopefully starting in Bachelor program by providing cheap solutions for practical work. The existence of such cheap material will pave the way for neuroscientific based games. Moreover preserving animal life is at the core of the “In vitro Artificial Intelligence” project. Cells in culture can be programmed into endless multiplying cells freeing neuroscientists of the regular need of fresh cells from living animals.
II Description of an electrophysiological amplifier
An “amplifier” mostly conveys electrical signal from one place - the electric signal obtained in each recording sites below the tissue- to another one -within the computer for storage, representation and handling- meanwhile increasing the size of the electric signal from micro volts (μV) to Volts (V). To perform such a task reliably, two features need to be simultaneously preserved: a μV-sized stable baseline and a reliable control of the signal over noise ratio through the amplifications. Usually, electrophysiological research material -including the open-source material from Backyard Brain, OpenBCI and Open-Ephys- succeed in performing these two tasks by manipulating the electric signal in two separate elements: the pre-amplifier and the amplifier. One of the amplifier’s role is to isolate the noise from the external world in a space where the preamplifier and the biological material will be placed. The most classical noise to get rid of is the standard electronic noise -50 or 60Hz- from common wall wire’s alternative current. Doing so, the amplifier provides a stable μV-sized electronic baseline, along with a noise-free power supply to the pre-amplifier. The pre-amplifier is composed of μV-sensitive components that transform the μV-sized signal observed inside the culture into V-sized signal for the amplifier.
Such hierarchal electric handling will be reproduced for the proven robustness of the electronic amplification it provides. The question whether or not to intake made-up solution from Backyard brain, OpenBCI or Open-Ephys to build similar solutions our self, is to be answered by the potential team, in agreement with a reasonable timeline and a proper budget validation. A minimal, over-simplified set-up would be:
a multi-electrode array, which can be either provided by the mentor (Multichannel system 64-channels Multi-electrode-Array) or made locally (cf. Fig. 1)
a pre-amplifier coupled to an amplifier, which can be either build locally (Giovangrandi Kovacs 2005), or intaken from OpenBCI. A third non-educational option would be to rent an amplifier to Multichannels system and then mimic it.
A digitizer, usually included in the amplifier. This element is made to perform the connection computer to amplifier. It is the location where real-time hardware-based analysis is usually performed (Zrenner Marom 2010).
Detailed options with timeline are listed in the following paragraph. The actual plan for this project should be very carefully prepared prior to any decisions, according to the chosen option and mostly to the composition of the student’s team potentially motivated in such projects.
III Timeline, material and potential versions
Manipulation of biological material is not part of this proposal and student contact with any biological agents will be avoided. The mentors can provide extensive dataset including in vivo, in vitro and ex vivo recordings, if necessary. The mentor guarantees his presence and takes complete responsibility of handling himself potential biological agents. In such unexpected outcome, the mentors could either provide biological materials, obtained under the appropriate conditions in partner laboratory in Paris, or simply invite temporarily the team and the material to a laboratory for proper manipulation of biological agents.
Several options are available towards prototypes of amplifiers. The most reasonable options would be to focus on each elements of the amplifier, one at a time. Four different options are here proposed:
The recording chamber option: built in-house ‘flex-MEAs’, from already published examples cf. Fig1 (Giovangrandi Kovacs 2005).
The amplifiers option: create the combined preamplifier and amplifier system, for multiple recording sites. Open-BCI freely provides their system architecture, which can be adapted or mimicked.
The digitizer option: focus on the hardware contained in the digitizer unit. The analog-digital conversion should be very easy to reproduce, here the question would be which other of the basic analysis functions could be implemented? The first wanted function is the filtering and extraction of action potentials. The last, but not least, would be to produce “neuronal response clamp” (Wallach Marom 2011, Zrenner Marom 2010).
The full scale option: reproduce a whole set-up including the different elements quoted before. One three weeks doable objective would be to directly adapt usual MEA onto an Open-BCI amplifier system, which already perform reliably all the basic functions we want. From there, mimicking the different components of this system will be an easier task.
The different alternatives proposed should help to better create either: a match between competences and a (piece of) prototype, or to invite students into fostering interest in doing parts of this project over a longer period of time.
IV Future directions
This project proposes to build an easy-to-use low cost open-source materials that could challenge the available industrial products, as done by OPEN-Ephys and OPEN-BCI since couple of years. The core interest of such a project is both open-knowledge for scientific research purposes and educative objectives through gaming.
Fig 1: Design of flex-MEAs (Giovangrandi Kovacs 2005): (a) CAD layout for top (electrode) and bottom (heater) layers; (b) cross-section schematics of the flex structure; © photograph of a flex-MEA as received from the manufacturer, with close-ups of the electrode area and of a 65μm electrode; (d) photograph of an assembled flex-MEA, mounted on a rigid substrate with connecting pins and attached to a cell culture dish. The two large electrodes (635 μm × 2800 μm) visible in the center of the circuit are used to ground the culture bath during recording. The bottom pins connects the recording electrodes, while the top pins connects the heater and temperature sense traces.