Neuron network modeling

• US 6,865,563 B1
• Filed: 05/09/2002
• Issued: 03/08/2005
• Est. Priority Date: 05/09/2001
• Status: Active Grant

• Current Assignee: Maktab Tarighat Oveyssi Shahmaghsoudi
• Original Assignee: Maktab Tarighe Oveyssi Shah Maghsoudi
• Inventors: Angha, Nader
• Primary Examiner(s): Davis, George B.
• Application Number:
  US10/141,069
• Time in Patent Office: 1,034 Days
• Field of Search: 706/15, 706/33
• US Class Current: 706/15
• CPC Class Codes: G06N 3/065 Analogue means
• CPC Class Codes: G10L 15/16 using artificial neural networks

Abstract

The Neuron network modeling compares neuron behavior in terms of electrical properties to predict the propagation of capacitative, or inductive current/potential along a nerve fiber. Membrane ion channels act as capacitor-inductor circuits, similar to a radio receiver and transmitter. Therefore, a single circuit, including the membrane capacitor, the ion channel, inductors and resistors, nerve fibers, neurons, and the whole CNS may be modeled as radio receivers and transmitters A speech process for the hearing impaired is disclosed as one example embodiment.

Claims

1. A neuron network modeling apparatus for the propagation of current potential along a selected nerve fiber, which comprises:

a neuron, two membranes, and membrane ion channels, which together act as a resistor-capacitor-inductor circuit to enhance the reception-transmission power of the neuron network in a neuron computer chip.

2. Neuron network modeling apparatus of claim 1, wherein membrane ion channels in the neuron network modeling apparatus act as inductive elements that conduct the ions on spiral paths.

3. The neuron network modeling apparatus of claim 1, wherein stimuli is used to increase the positive ion concentration of the outer cell membrane, in order to increase the magnitude of the potential depolarization of the neuron network modeling apparatus.

4. The neuron computer chip apparatus of claim 1, wherein all nerve fibers between the point of stimulation and the data processing center are set to oscillate at the same frequency within the neuron network modeling apparatus.

5. The neuron network modeling apparatus of claim 1, wherein the neuron network modeling apparatus communicates with other neuron networks within the neuron network modeling apparatus without requiring any physical contact there-between.

6. The neuron network modeling apparatus of claim 1, wherein a neuron computer chip converts speech to text.

7. The neuron network modeling apparatus of claim 1, wherein a membrane current is positioned perpendicular to a selected membrane, and the internal current is directed along the axis of a cylindrical nerve fiber located in the neuron network modeling apparatus.

8. The neuron network modeling apparatus of claim 1, wherein neuron behavior is delivered in terms of electrical properties pertaining to the propagation of resistive, capacitative and inductive current/potential along a nerve fiber within the neuron network modeling apparatus.

9. The neuron network modeling apparatus of claim 1, wherein the two membranes act as capacitors and the ion channels act as an inductor that conduct the ions on spiral paths in an electrical circuit, in a manner similar to a radio receiver and transmitter within the neuron network modeling apparatus.

10. A process of neuron network modeling for the propagation of current potential along a selected nerve fiber, which comprises:

combining a neuron, two membranes, and at least one membrane ion channel, which together act as a resistor-capacitor-inductor circuit to enhance the reception-transmission power of the neuron network in a neuron computer chip.

11. The process of neuron network modeling of claim 10, wherein the membrane ion channels act as inductive elements that conduct the ions on spiral paths within the neuron network.

12. The process of neuron network modeling of claim 10, wherein stimuli is used to increase the positive ion concentration of the outer cell membrane, in order to increase the magnitude of the potential depolarization within the neuron network.

13. The process of neuron network modeling of claim 10, wherein all nerve fibers between the point of stimulation and the data processing center are set to oscillate at the same frequency within the neuron network.

14. The process of neuron network modeling of claim 10, wherein a voice input is sent to a preprocessor, and a signal from the preprocessor then goes to a speech processor IC comprising a network of neurons, the speech processor IC in communication with a database of words and phrases in at least one language, and the speech processor IC sends a print relating to the voice input to a display monitor for converting speech to text.

15. The process of neuron network modeling of claim 10, wherein a membrane current is positioned perpendicular to a selected membrane, and the internal current is directed along the axis of the cylindrical nerve fiber within the neuron network.

16. The process of neuron network modeling of claim 10, wherein neuron behavior is delivered in terms of electrical properties pertaining to the propagation of resistive, capacitative and inductive current/potential along a nerve fiber within the neuron network.

17. The process of neuron network modeling of claim 10, wherein the two membranes act as capacitors and the ion channels act as an inductor having a spiral path in an electrical circuit, in a-manner similar to a radio receiver, a transmitter.

18. A method of neuron network modeling, which comprises:

a) processing an input voice signal in a preprocessor;

b) sending the signal from the preprocessor to a network of neurons forming a speech processor integrated circuit;

c) accessing a database of words and phrases in at least one selected language with the speech processor; and

d) and visually displaying a print of the input voice on a display monitor.

19. The method of claim 18, which comprises increasing the positive ion concentration of the outer cell membrane, thus increasing the magnitude of the potential depolarization in the network of neurons.

20. The method of claim 18, which comprises delivering neuron behavior in terms of electrical properties pertaining to the propagation of resistive, capacitative and inductive current/potential along a nerve fiber.