Apparatus for treating a patient

Abstract

At least one embodiment of the disclosure is directed to a method for processing brainwave signals for treating a patient having neurological disorder or mental disorder or a combination of neurological and mental disorder. The method comprises: measuring a brainwave signal from the patient, the measured brainwave signal containing noise; denoising the brainwave signal to obtain a clean brainwave signal; matching the clean brainwave signal to a database of brainwave signals for neurological or mental conditions or a combination of neurological and mental conditions to identify the patient's neurological or mental status or a combination of neurological and mental conditions; and applying a therapeutic treatment to the patient based on the identified neurological or mental status or a combination of neurological and mental conditions.

Description

CROSS-REFERENCE TO RELATED APPLICATION[0001]This application is a continuation application of U.S. application Ser. No. 12 / 804,936, filed Aug. 2, 2010; the contents of the above application are incorporated by reference in its entirety.I. FIELD OF THE INVENTION[0002]The present invention relates to signal processing, and more particularly to a signal processing method and system for the extraction of signals of unknown or unspecified form or features with severe noise corruption, such as MEG or EEG signals. The invention also relates to generating desired responses to the detected clean signals, in particular to medical diagnostic and treatment methods and apparatus, and more particularly to detection and treatment of mental / neurological conditions.II. BACKGROUND OF THE INVENTION[0003]Sensors or instrumentation deployed in real-world settings for various fields, e.g. analytical medicine, and using various detection modalities, e.g. electroencephalography, sonography, and electrocard...

AI Technical Summary

Benefits of technology

[0015]Another aspect of the present invention includes a system comprising: a discrete wavelet transformer capable of iteratively decomposing a signal into a plurality of decomposition levels, each having a smooth component and a rough component; at least one auto-associative neural network, each auto-associative neural network operatively coupled to the discrete wavelet transformer to receive a corresponding pre-selected component, and to squeeze out noise in the wavelet domain from a corresponding pre-selected component through constraining the number of neurons in the hidden layer; and an inverse discrete wavelet transformer capable of recovering a clean signal in the time domain from the combined outputs of the plurality of neural networks.

Problems solved by technology

Sensors or instrumentation deployed in real-world settings for various fields, e.g. analytical medicine, and using various detection modalities, e.g. electroencephalography, sonography, and electrocardiography, usually produce signals corrupted by various types of noise.

For example, noisy data would make data compression much harder and thus affect storage and transmission, e.g. storing X-rays and mammograms.

If not done correctly, the typical preprocessing inherent in any instrument design can actually remove valuable information, and subsequent use of even advanced signal processing methods, no matter how capable, will not be able to recover signal lost by inadequate pre-processing.

However, there are signals that are of unknown or unspecific nature or shapes because we do not know the theoretical expected behavior.

Magnetoencephalography or MEG and Electroencephalography or EEG signals from the brain are highly complex and at this junction our understanding of the brain is incomplete and thus we do not have a theoretical expected signal behavior from the brain.

As a result, any neural network based on supervised training is useless since we really do not know the real target signal for it cannot be derived from first principle.

All drugs have undesirable side effects and relative risks.

However, there is no consistent information about whether these interventions are helpful in improving other domains of impairment and associated disability, even though these problems are often the greatest concern to patients; nor does the available evidence specify when, and for whom, various psychotherapeutic interventions should be provided, or whether different treatment modalities can and should be combined, or how they should be combined.

However, very little research has been pursued using dual frequencies (low and high), possibly because the size of the coil makes it hard to have two coils operating simultaneously.

For most current Transcranial Magnetic Stimulation (TMS) work, the double-loop coil used measures 174 mm—almost the size of a human head, which may make it difficult to localize the stimulated region of the brain.

Images