Electroencephalography monitoring device having a self-adaptive analog-to-digital converter

1. An electroencephalography (EEG) monitoring device, comprising:

at least one electrode;

a non-linear analog signal processor configured to receive a signal from the at least one electrode and determine whether the signal is within a significant category, wherein a significant EEG event is within the significant category, the non-linear analog signal processor comprising:

an independent component analyzer configured to provide blind source separation functionality at least in part by implementing an adaptation rule as a product of a sign of a first processed signal from a first electrode, a square of the first processed signal, and a hyperbolic tangent of a second processed signal; and

a feature extractor for extracting representative statistical features of an EEG signal configured to indicate occurrence of a seizure event;

a variable resolution analog-to-digital converter configured to convert signals received from the non-linear analog signal processor into high-resolution digital signals if the signals are within the significant category, and configured to convert signals received from the non-linear analog signal processor at a low resolution if the signals are not within the significant category;

storage means configured to store high-resolution digital signals received from the variable resolution analog-to-digital converter; and

a power source connected to, and providing power to, the non-linear analog signal processor, the variable resolution analog-to-digital converter, and the storage means.

2. The EEG monitoring device of claim 1 further comprising at least one amplifier connected to the at least one electrode and configured to amplify a signal received from the at least one electrode.

3. The EEG monitoring device of claim 1 , further comprising RF modulation circuitry and a driver configured to receive at least one of a high-resolution digital signal and a low-resolution digital signal from the variable resolution analog-to-digital converter.

4. The EEG monitoring device of claim 3 , further comprising an antenna connected to said driver.

5. The EEG monitoring device of claim 1 , wherein a signal within the significant category represents an ictal episode.

6. The EEG monitoring device of claim 1 , wherein the low-resolution digital samples are not stored within the storage means.

7. The EEG monitoring device of claim 1 , wherein the storage means is a memory.

8. The EEG monitoring device of claim 1 , wherein the storage means is a memory located within a microprocessor.

9. The EEG monitoring device of claim 8 , wherein the memory is configured to store the high-resolution digital signals until filled, after which the microprocessor is configured to transmit the high-resolution digital signals wirelessly.

10. The EEG monitoring device of claim 1 , wherein the power source is a removable battery.

11. The EEG monitoring device of claim 1 , wherein the non-linear analog signal processor further comprises

an event classifier configured to be trained either on patient-specific data, or on data of a patient population.

12. The EEG monitoring device of claim 1 , wherein the independent component analyzer comprises a squaring circuit comprising a V-I square law of a transistor.

13. The EEG monitoring device of claim 12 , wherein an output of the squaring circuit is used as a bias current of an operational-transconductance amplifier whose non-inverting input is the second signal.

14. The EEG monitoring device of claim 1 , wherein the independent component analyzer comprises a switch configured to turn on or off depending on the sign of the first signal.

15. An electroencephalography (EEG) monitoring device, comprising:

at least one electrode;

a non-linear analog signal processor configured to receive a signal from the at least one electrode and determine whether the signal is within a significant category, wherein a significant EEG event is within the significant category, the non-linear analog signal processor comprising:

an independent component analyzer configured to provide blind source separation functionality; and

a feature extractor for extracting representative statistical features of an EEG signal configured to indicate occurrence of a seizure event at least in part by

applying a peak value of an input signal to a first transconductance element and a second transconductance element, and

with a multiplexer, multiplexing the outputs of the first transconductance element and the second transconductance element according to a comparator configured to compare the input signal to the output signal,

wherein the output signal comprises the output of the multiplexer.

a variable resolution analog-to-digital converter configured to convert signals received from the non-linear analog signal processor into high-resolution digital signals if the signals are within the significant category, and configured to convert signals received from the non-linear analog signal processor at a low resolution if the signals are not within the significant category;

storage means configured to store high-resolution digital signals received from the variable resolution analog-to-digital converter; and

a power source connected to, and providing power to, the non-linear analog signal processor, the variable resolution analog-to-digital converter, and the storage means.

16. The EEG monitoring device of claim 15 , wherein the comparator is configured to approximate a heavyside operator.

17. The EEG monitoring device of claim 15 , wherein the peak value of the input signal comprises the output of circuitry comprising a half-wave rectifier and a capacitor.

18. The EEG monitoring device of claim 15 , wherein the input signal comprises an output of the independent component analyzer.

19. The EEG monitoring device of claim 15 , wherein the multiplexer comprises a current-mode multiplexer.