Transcript finkler_20070925.ppt

Silicon Chip Interfaced with a
Geometrically Defined Net of
Snail Neurons
Authors: M. Merz, and P. Fromherz
Presenter: Todd Finkler
Problem

Analyzing real neural networks
◦ Inaccessible to non-invasively excite
◦ Currently impossible to record single cell
resolution

So…create low density cultures
◦ Drastically reduce complexity, BUT…
◦ Synaptic wiring still unclear
Goal

Build neural network with defined
topology
Proposed Methodology
Mechanical
immobilization
 Topographical
guidance of neurites

Polyester film over silicon chip
(neurons cultured 2 days)
Proposed Methodology (cont.)

Electrical interface
of individual
neurons
◦ Open-gate FET –
recording
◦ Capacitor –
stimulation

Control – impaled
microelectrode
Selected Neurons Cultured

Pond snail neurons (Lymnaea stagnalis)
◦
◦
◦
◦
Large size
High extracellular currents
Good SNR in transistor recording
Active in low-density cultures
Single Neuron
Induced Action Potential
Ten square wave pulses
•2 V amplitude
•1 ms duration
Microelectrode
(intracellular)
Transistor
(extracellular)
16 mV
Neuron Pair
Intracellular
(Microelectrode)
1
1
2
2
Intracellular
(Microelectrode)
Extracellular
(Transistor)
Neural Network (4 cells)
Wiring
Optical
Microscope
Fluorescence
Microscope
(1,3,4 stained)
Neural Network (4 cells)
Spontaneous firing
Measurement
Capacitor stimulated
1,2 & 3 Impaled
microelectrode
Transistor
Capacitor stimulated
Cultivating the Network

Not high enough yield for a complex
hybrid system
◦
◦
◦
◦
Only 23% of 4000 neurons grew neurites
Only 62% of above stayed in the tunnels
Of those 62%, only 29% achived bi-directional
Total yield – 4%
(920)
(570)
(165)
Questions