Method for the analysis, display and classification of multivariate indices of brain function--a functional electrophysiological brain scan

• US 4,188,956 A
• Filed: 06/26/1978
• Issued: 02/19/1980
• Est. Priority Date: 06/26/1978
• Status: Expired due to Term

• Current Assignee: New York Medical College (Touro College & University System)
• Original Assignee: John E. Roy
• Inventors: John, E. Roy
• Primary Examiner(s): Kamm, William E.
• Application Number:
  US05/918,730
• Time in Patent Office: 603 Days
• Field of Search: 128/2.1 B, 128/731
• US Class Current: 600/544
• CPC Class Codes: A61B 5/276 Protection against electrode failure
• CPC Class Codes: A61B 5/291 for electroencephalography [EEG]
• CPC Class Codes: A61B 5/377 using evoked responses
• CPC Class Codes: A61B 5/7232 involving compression of the physiological signal, e.g. to extend the signal recording period
• CPC Class Codes: A61B 5/7264 Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems neural networks per se G06N3/00; expert systems per se G06N5/00
• CPC Class Codes: G16H 50/20 for computer-aided diagnosis, e.g. based on medical expert systems

Abstract

A method in electroencephalography (EEG) for the display of neurometric test data in which electrical activity generated by the brain either spontaneously or in response to specified challenges or conditions is detected at specified locations on the head surface of the subject by a plurality of electrodes. Features extracted from the spontaneous or evoked brain waves are quantified according to specified neurometric indices and categorized, using the mean values and standard deviations of control groups, according to their relative probability in a population of normal healthy people. Multivariate methods such as multiple analysis of variance or stepwise discriminate analysis can be used to identify independent features particularly useful for diagnosis. The values of such features can be used to construct a multivariate vector for each individual patient. If the length of this vector for data from a specified electrode placement is above the norm by a predetermined amount, for example, two standard errors, a plus sign is plotted at the corresponding position in a head diagram. If it is below the norm, for example, by more than two standard deviations (p<0.01), a minus sign is plotted on the head diagram.

Claims

1. A method for the acquisition, compression, analysis and display of neurometric test data which comprises

connecting a plurality of electrically responsive terminals to the subject'"'"'s scalp,

exposing the subject to a specified set of conditions to evoke the subject'"'"'s brain wave responses reflecting a variety of types of brain functions,

sensing the subject'"'"'s brain electrical responses,

statistically analyzing said responses using a digital computer based system,

representing the results of said statistical analysis as hard copy display according to a locational sign character convention in which the representation occurs at locations on the copy corresponding in format to the locations of said electrodes on the subject'"'"'s head, and the deviations of said results above a statistically predetermined normal range of values are plotted as plus signs and values below the predetermined range of values are plotted as minus signs, with the size and density of the plotted symbols proportional to the statistical significance of such deviations.

2. A method as in claim 1 wherein values within the predetermined normal range are plotted as two diagonalized dots.

3. A method as in claim 1 wherein the data are statistically compressed by means selected from the group of F-ratio, multiple analysis of variance, factor analysis, and stepwise discriminant analysis.

4. A method as in claim 1 wherein the data is statistically evaluated by the Z-distribution.

5. A method as in claim 1 wherein values greatly above or below the predetermined range are encoded respectively as plus and minus signs whose size and density correspond in proportion to the significance of the deviation.

6. A method as in claim 1 and including the further steps of plotting the values for the subject across a set of conditions or challenges in multidimensional space to construct a vector indicative of the subject and grouping closely situated vectors together to define a taxonomic class when the distance between the vectors is less than the distance between individual vector points.

8. A method as in claim 1 wherein values within the predetermined range are plotted as two diagonalized dots.

9. A method as in claim 1 wherein the data are compressed by use of multivariate statistical procedures.

7. A method for the compression, analysis and display of neurometric test data which comprises

connecting a plurality of electrically responsive terminals to the subject'"'"'s scalp in a predetermined spacial format,

sensing the subject'"'"'s brain electrical activity,

statistically analyzing said activity using a digital computer based system as compared to the activity of other subjects under similar conditions,

representing the results of said statistical analysis as a hard copy display according to a locational sign character conventional in which the format resembles the head of the subject as seen from above and at a plurality of locations within said head-shape, the values of said results above a statistically determined range of values are plotted as plus signs and values below the statistically determined range of values are plotted as minus signs.