METHOD AND SYSTEM FOR MOTOR REHABILITATION

• US 20140018694A1
• Filed: 03/31/2011
• Published: 01/16/2014
• Est. Priority Date: 03/31/2010
• Status: Active Grant

• Current Assignee: Agency For Science Technology and Research, Tan Tock Seng Hospital Pte Limited (MOH Holdings Pte Limited)
• Original Assignee: Agency For Science Technology and Research, National Health-Care Group Pte Limited
• Inventors: Ang, Kai Keng, Guan, Cuntai, Wang, Chuanchu, Phua, Kok Soon, Zhang, Haihong, Chin, Zheng Yang, Chua, Sui Geok Karen, Tan, Hock Guan Adrian
• Granted Patent: US 9,357,938 B2
• Time in Patent Office: Days
• US Class Current: 600/544
• CPC Class Codes: A61B 5/316 Modalities, i.e. specific diagnostic methods
• CPC Class Codes: A61B 5/374 Detecting the frequency distribution of signals, e.g. detecting delta, theta, alpha, beta or gamma waves
• CPC Class Codes: A61B 5/375 using biofeedback
• CPC Class Codes: A61B 5/4205 Evaluating swallowing
• CPC Class Codes: A61B 5/7267 involving training the classification device
• CPC Class Codes: A61N 1/36003 of motor muscles, e.g. for walking assistance
• CPC Class Codes: G16H 50/70 for mining of medical data, e.g. analysing previous cases of other patients

Abstract

A method of calibrating a motor imagery detection module and a system for motor rehabilitation are provided. The method comprises acquiring Electroencephalography (EEG) data from a subject; selecting classification features from the EEG data; wherein the feature selection comprises modelling an idle state ω_n of the subject by M sub-classes χ_j, j=1, . . . , M.

Claims

1. A method of calibrating a motor imagery detection module, the method comprising,

acquiring Electroencephalography (EEG) data from a subject;

decomposing the acquired EEG data into a plurality of frequency and time segment components;

computing a projection matrix for spatial filtering for each of the plurality of frequency and time segment components;

selecting classification features from the plurality of frequency and time segment components for calibrating the motor imagery detection module;

wherein the feature selection comprises modelling an idle state ω

_n of the subject by M sub-classes χ

_j, j=1, . . . , .

3. The method as claimed in claim 1, wherein selecting the classification features further comprises computing candidate features of each of the plurality of frequency and time segment components.

4. The method as claimed in claim 3, wherein selecting the classification features comprises computing mutual information of each candidate feature for each of the plurality of frequency and time segment components.

5. The method as claimed in claim 4, wherein selecting the classification features comprises selecting, for each of the plurality of frequency and time segment components, the candidate features with maximum mutual information with motor imagery actions and rest respectively.

6. The method as claimed in claim 5, further comprising training a non-linear regression model using the selected features.

17. A computer readable data storage medium having stored thereon computer code means for instructing a computer processor to execute a method of calibrating a motor imagery detection module as claimed in claim 1.

2. (canceled)

7. (canceled)

8. A method of motor rehabilitation, the method comprising,

extracting features from Electroencephalography (EEG) data of a subject, wherein the feature extraction comprises using a multi-modal model for an idle state ω

_n of the subject including M sub-classes χ

_j, j=1, . . . , M;

using a rehabilitation module to monitor an output of the multi-modal model for motor control signals;

and if motor control signals are detected,

applying functional electrical stimulation (FES) to the subject.

9. A system for motor imagery detection, the system comprising,

a signal amplifier capable of acquiring Electroencephalography (EEG) data from a subject, and

a detection module for detecting motor imagery based on the EEG data received from the signal amplifier;

wherein the detection module is configured to decompose the acquired EEG data into a plurality of frequency and time segment components, compute a projection matrix for spatial filtering for each of the plurality of frequency and time segment components, and select classification features from the plurality of frequency and time segment components; and

wherein the feature selection comprises modelling an idle state ω

_n of the subject by M sub-classes χ

_j, j=1, . . . , M.

11. The system as claimed in claim 9, wherein selecting the classification features comprises computing mutual information of each candidate feature for each of the plurality of frequency and time segment components.

13. The system as claimed in claim 11, wherein selecting the classification features comprises selecting, for each of the plurality of frequency and time segment components, the candidate features with maximum mutual information with motor imagery actions and rest respectively.

14. The system as claimed in claim 13, wherein the detection module is further configured to train a non-linear regression model using the selected features.

10. (canceled)

12. (canceled)

15. (canceled)

16. (canceled)

18. (canceled)

19. A method for brain-computer interface (BCI) based interaction, the method comprising the steps of:

acquiring a subject'"'"'s EEG data;

processing the EEG data to determine a motor imagery of the subject;

detecting a swallow movement of the subject using a detection device; and

providing feedback to the subject based on the motor imagery, the movement, or both;

wherein providing the feedback comprises activating a stimulation element for providing a stimulus to throat muscles of the subject, and

wherein the processing of the EEG data comprises selecting classification features from the EEG data, and wherein the feature selection comprises modelling an idle state ω

_n of the subject by M sub-classes χ

_j, j=1, . . . , M.

20. The method as claimed in claim 19, comprising:

processing the EEG data to determine whether a specific swallow motor imagery is performed by the subject; and

activating the stimulation element if the specific swallow motor imagery is performed.

21. The method as claimed in claim 20, wherein the feedback further comprises a separate visual feedback to the subject if the specific swallow motor imagery is performed.

25. The method as claimed in claim 19, wherein the processing of the EEG data comprises using a trained classification algorithm for said selecting the features from the EEG data.

26. The method as claimed in claim 25, wherein training the classification algorithm comprises:

dividing the EEG data into a plurality of segments,

for each segment, dividing a corresponding EEG signal portion into a plurality of frequency bands,

for each frequency band, computing a spatial filtering projection matrix based on a CSP algorithm and a corresponding feature, and computing mutual information of each corresponding feature with respect to one or more motor imagery classes;

for each segment, summing the mutual information of all the corresponding features with respect to the respective classes, and

selecting the corresponding features of the segment with a maximum sum of mutual information for one class for training.

27. The method as claimed in claim 26, further comprising training classifiers of the classification algorithm using the selected corresponding features.

28. The method as claimed in claim 27, wherein training the classifiers comprises non-linear regression using the selected corresponding features and non-linear post-processing regression using an output from the non linear regression.

29. The method as claimed in claim 26, wherein computing the spatial filtering projection matrix based on the CSP algorithm comprises using a multi-modal multi-time segment for each frequency band.

30. The method as claimed in claim 29, wherein the multi-modal multi-time segment for each frequency band comprises a multi-modal representation of idle states.

22. (canceled)

23. (canceled)

24. (canceled)

31. (canceled)

32. (canceled)

33. (canceled)

34. (canceled)