U.S. patent application number 09/930632 was filed with the patent office on 2003-07-31 for eeg prediction method for medication response.
Invention is credited to Emory, W. Hamlin, Suffin, Stephen C..
Application Number | 20030144875 09/930632 |
Document ID | / |
Family ID | 46280055 |
Filed Date | 2003-07-31 |
United States Patent
Application |
20030144875 |
Kind Code |
A1 |
Suffin, Stephen C. ; et
al. |
July 31, 2003 |
EEG prediction method for medication response
Abstract
The present invention includes a system and method for
computerized analysis of a patient's electroencephalogram (EEG)
recorded by electrodes placed on the scalp, for the purpose of
predicting patient response to medications and therapeutic agents
commonly used in psychiatric practice. The prediction of the
responses to medications (adverse, no effect, favorable outcome) is
an important problem in the clinical practice of psychiatry. A
growing number of therapeutic agents are available to the clinician
but these agents generate variable responses when prescribed based
solely on the patient's history and current symptoms. The present
invention is used by physicians to improve patient outcome by
selecting agents most likely to be effective for a given patient,
using a standardized analysis of the digitized EEG and comparison
of individual patient EEC data to a particular database of similar
patients whose clinical outcome to pharmacotherapy is known.
Inventors: |
Suffin, Stephen C.; (Sherman
Oaks, CA) ; Emory, W. Hamlin; (Malibu, CA) |
Correspondence
Address: |
PENNIE AND EDMONDS
1155 AVENUE OF THE AMERICAS
NEW YORK
NY
100362711
|
Family ID: |
46280055 |
Appl. No.: |
09/930632 |
Filed: |
August 15, 2001 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
09930632 |
Aug 15, 2001 |
|
|
|
09148591 |
Sep 4, 1998 |
|
|
|
60058052 |
Sep 6, 1997 |
|
|
|
Current U.S.
Class: |
705/2 |
Current CPC
Class: |
A61B 5/374 20210101;
G16H 80/00 20180101; A61B 5/165 20130101; G16H 50/20 20180101; G16H
40/67 20180101; A61B 5/16 20130101; G06F 21/6254 20130101 |
Class at
Publication: |
705/2 |
International
Class: |
G06F 017/60 |
Claims
We claim:
1. A unique system for compressing, encrypting, tracking, and
securely transmitting digital EEG data and associated patient
identifying information via the Internet from a remote site to a
Report Processing Center, and including the electronic return of a
report summarizing results of proprietary analyses and database
comparison all without requiring telephonic transmission.
2. Identification of a set of univariate and multivariate EEG
features that when observed in a patient diagnosed with a
psychiatric disorder, can be used with NuPharm Database's
particular rule-based classifier to predict a favorable clinical
responsive to psychostimulant class medications.
3. Identification of a set of univariate and multivariate EEG
features that when observed in a patient diagnosed with a
psychiatric disorder, can be used with NuPharm Database's
particular rule-based classifier to predict a favorable clinical
responsive to antidepressant class medications.
4. Identification of a set of univariate and multivariate EEG
features that when observed in a patient diagnosed with a
psychiatric disorder, can be used with NuPharm Database's
particular rule-based classifier to predict a favorable clinical
response to anticonvulsant class medications.
5. Identification of a set of univariate and multivariate EEG
features that when observed in a patient diagnosed with a
psychiatric disorder, can be used with NuPharm Database's
particular rule-based classifier to predict a favorable clinical
responsive to a combination of psychostimulant and antidepressant
class medications.
6. Identification of a set of univariate and multivariate EEG
features that when observed in a patient diagnosed with a
psychiatric disorder, can be used with NuPharm Database's
particular rule-based classifier to predict a favorable clinical
responsive to a combination of anticonvulsant and antidepressant
class medications.
7. Identification of a set of univariate and multivariate EEG
features that when observed in a patient diagnosed with a
psychiatric disorder, can be used with NuPharm Database's
particular rule-based classifier to predict a favorable clinical
response to a combination of psychostimulant, antidepressant, and
anticonvulsant class medications.
8. A method for computerized generation of clinical reports that
integrates interpretive information from medical professionals with
results of medication responsivity evaluation according to claim 2.
Description
[0001] antidepressant class medications, anticonvulsant class
medications, combinations of psychostimulant and antidepressant
class medications, combinations of anticonvulsant and
antidepressant class medications, combinations of psychostimulant,
antidepressant, and anticonvulsant class medications.
[0002] The present invention also includes a method for
computerized generation of clinical reports that integrates
interpretive information from medical professionals with results of
medication responsivity evaluation.
BRIEF DESCRIPTION OF THE FIGURES
[0003] The present invention may be understood more fully by
reference to the following detailed description of the preferred
embodiment of the present invention, illustrative examples of
specific embodiments of the invention and the appended figures in
which
[0004] FIG. 1 illustrates a method of the present invention where:
step 1 of FIG. 1 corresponds to elements 1 and 2 of the invention
described below; step 2 corresponds to elements 3, 4, and 3; step 3
to elements 6 and 7; step 4 to element 8; and step 5 to elements 9
and 10.
DETAILED DESCRIPTION OF THE INVENTION
[0005] More specifically, the following steps are employed:
[0006] 1) The EEG is recorded using electrodes placed on the
patient's scalp, and the EEG data is stored in a digital format
using a standardized protocol available on one of a number of
commercially available instruments (current manufacturers include
Cadwell Laboratories, Bio-Logic Systems Corp., Nicolet Biomedical,
Oxford Instruments, among others). The International 10-20 System
convention is used for determining the location of electrodes
placed on the scalp. It is the responsibility of the recording
facility to collect data in accordance with procedural
specifications.
[0007] 2) The following patient criteria apply:
[0008] a) Patient must have received a psychiatric diagnosis as
specified in the Diagnostic and Statistical Manual, currently the
Fourth Edition (DSM-IV).
[0009] b) Ages between six and ninety.
[0010] c) Patient is taking no medications. All medications
potentially influence the EEG and must be discontinued or avoided
for seven half-lives prior to baseline EBG examination. This
includes "over the counter" sleeping pills, pain medication,
nutritional health supplements and mega-vitamins.
[0011] d) Insulin, thyroid, estrogen, progesterone and other
hormone replacement agents are not excluded. Some cardiac agents
are included in the reference population of after the age of
fifty-five.
[0012] e) Patients with any of the characteristics listed below are
not suitable for prediction of medication responsivity based on EEG
analysis:
[0013] (i) intramuscular depo-neuroleptic therapy within the
preceding twelve months
[0014] (ii) a history of craniotomy with or without metal
prostheses
[0015] (iii) a history of cerebrovascular accident
[0016] (iv) spikes or extreme low voltage on the conventional
EEG
[0017] (v) a current diagnosis of seizure disorder
[0018] (vi) a diagnosis of dementia
[0019] (vii) mental retardation
[0020] (viii) current use of marijuana, cocaine, hallucinogens or
other drugs of abuse
[0021] (ix) inability to remain medication-free and drug-free for
seven half-lives of the current agent(s) prior to EEG recording
[0022] (x) significant abnormality of the CBC, chemistry or thyroid
panel with TSH until corrected
[0023] f) A "positive" Urine Drug Screen (UDS) interferes with
medication prediction methods. Studies are processed only if the
UDS is negative just prior to recording the digital EEG.
[0024] 3) The digital EEG data computer file is packaged along with
additional patient identifying information using packaging and
transmission software. The patient information includes:
[0025] a) name
[0026] b) date of birth
[0027] c) referring physician
[0028] d) handedness
[0029] e) height
[0030] f) weight
[0031] g) date of test
[0032] h) patient ID (social security number)
[0033] Packaging refers to compression of the computer file and
encryption of the file so that it cannot be opened or examined by
anyone other than at the processing center. The data transfer is
rigorously secured to protect the confidentiality of patient
records. The EEG files are encrypted at the recording facility with
a key known only to processing center. The patient ID is
transformed using a algorithm so that even in the case of mail
routing error there is no way to associate the data with an
individual. The data is compressed and protected with an additional
password and data files are transmitted to a secure site. These
steps mean that the patient data are protected against even
purposeful attempts to intercept and read them.
[0034] The transmittal of the EEG file and related patient
information is tracked as it is packaged, sent, processed, and
returned. All log entries include dates and times calibrated to
GMT.
[0035] The computer operating system preferred to run the packaging
and report transmission software is currently Microsoft Windows
95/98. The following hardware and software is preferred:
[0036] Hardware Requirements
[0037] Operating System: Windows 95 or Windows 98
[0038] Processor: 486, 133 MHZ.
[0039] Monitor and Video Card capable of displaying 256 colors.
[0040] Disk Space: 35 MB
[0041] RAM: 16 MB
[0042] CD-ROM Drive if installing from CD-ROM
[0043] Modem: 33.6 KBaud
[0044] Internet Connection with approved Internet Service
Provider
[0045] Software Requirements
[0046] Adobe Acrobat Reader Version 3.01
[0047] Microsoft Internet Explorer 4.0 or above
[0048] The packaging and transmission software
[0049] 4) The computer file is transferred off-hours using standard
commercially available file transfer protocols (FTP) via the
Internet, to a designated processing site. A special feature of the
packaging and transmission software exists to allow immediate
transfer of files for priority reporting if requested. The
processing site monitors the transfer in order to detect the
arrival of new computer files. When a new file is received, it is
forwarded for professional interpretation, if requested, and
specialized report generation.
[0050] 5) The file is decompressed and decrypted at the processing
site. Experienced technical and professional personnel then review
the EEG signals and sections of the recording identified as
containing signals generated by extracerebral sources are deleted
from subsequent analyses. The samples of EEG selected for inclusion
in analysis are then passed to the first stage of analysis.
[0051] 6) The first stage of analysis includes computations that
extract a standard set of features from the EEG. Quantitative
spectral analysis provides commonly used measures of EFG power and
relative power. Power is the square of amplitude; amplitude units
are in microvolts (.mu.V), power units are microvolts
squared(.mu.V.sup.2). Relative power is a measure of the proportion
of power in a given frequency band compared to the total band power
at a given electrode. Frequency bands are defined as delta, 0.5-2.5
Hz.; theta, 2.5-7.5 Hz.; alpha, 7.5-12.5 Hz., and beta, 12.5-32 Hz.
The total band is 0.5 to 32 Hz.
[0052] EEG coherence, a commonly used measure of the similarity of
activity for a pair of two scalp electrodes, also is extracted by
spectral analysis for all interhemispheric and intrahemispheric
sets of electrode pairs, for each frequency band as defined
above.
[0053] Commonly used measures of peak frequency within each defined
frequency band are computed.
[0054] Combinations of power and coherence measures over defined
sets of scalp electrodes are also computed.
[0055] 7) Features extracted from individual EEG data by
quantitative spectral and statistical analysis are further compared
to two distinct databases. In the second stage of analysis,
Z-scores representing deviations from a nonsymptomatic reference
population are computed. This reference population, often referred
to as the "Neurometric" database, contains 2082 quantitative EEG
measures including absolute power, relative power, coherence,
symmetry, and mean frequency of the delta, theta, alpha and beta
frequency bands of the EEG at every electrode position of the
International 10-20 System for individuals from 6 to 92 years
(database #1). The z-score value obtained by comparison of
individual's data to the age appropriate subset of the database
represents the patient's statistical deviation from the reference
database.
[0056] 8) The third stage of processing involves medication
response prediction using the patient database(database #2). This
prediction is made by first identifying the pattern of EEG
deviations from the reference database. Individual patient
deviation is then compared with the characteristic features of the
population of patients whose medications and treatment outcomes are
known. A rule-based classifier is applied to estimate the
likelihood that a patient EEG contains a pattern known to be
responsive to a given agent, class of agents, or combination of
agents or classes of agents. The EEG variables currently used by
the classifier are shown in Tables 1-4, below.
1 Column Column Heading Description of Abbreviation Heading
Description of Abbreviation Table 1 Table 2 RMAD Relative power
monopolar FMAD Frequency monopolar anterior delta anterior delta
RMPD posterior data FMPD posterior delta RMAT anterior theta FMAT
anterior theta RMPT posterior theta FMPT posterior theta RMAA
Anterior alpha FMAA anterior alpha RMPA Posterior alpha FMPA
posterior alpha RMAB Anterior beta FMAB anterior beta RMPB
posterior beta FMPB posterior beta CEAD Coherence interhemispheric
AADL Asymmetry intrahemispheric anterior delta delta - left CEPD
Posterior delta AADR delta - right CEAT anterior theta AATL theta -
left CEPT posterior theta AATR theta - right CEAA anterior alpha
AAAL alpha - left CEPA Posterior alpha AAAR alpha - right CEAB
Anterior beta AABL beta - left CEPB posterior beta AABR beta -
right Table 3 Table 4 AED Asymmetry monopolar CEBD Coherence
interhemispheric bipolar interhemispheric delta delta AFT Theta
CEBT Theta AEA Alpha CEBA Alpha AEB Beta CEBB Beta AEBD Asymmetry
bipolar RBDL Relative power bipolar delta left interhemispheric
delta AEBT Theta RBDR Delta - right AEBA Alpha RBTL Theta - left
AEBB Beta RBTR Theta - right CADL Coherence intrahemispheric RBAL
Alpha - left delta - left CADR Delta - right RBAR Alpha - right
CATL Theta - left RBBL Beta - left CATR Theta - right RBBR Beta -
right CAAL Alpha - left CAAR Alpha - right CABL Beta - left CABR
Beta - right
[0057] 9) A formal report for the referring clinician is generated.
The report is returned in a format that cannot be modified by the
client (Adobe Systems, Inc., "portable document format", or "PDF").
This report contains certain elements as specifically requested by
the referring clinician. These elements may include a professional
medical interpretation of the digital EEG tracing, a presentation
of selected features extracted by quantitative EEG analysis, a
presentation of deviations from the Neurometric database, and a
statement of the likelihood of favorable pharmacotherapeutic
outcome based on comparison with patients having similar EEG
features in the patient database #2. The treating physician is
responsible for any medication selection, titrating of dosage and
monitoring the patient for side effects and is instructed to
incorporate results of reports with the psychiatric assessment to
develop into an overall clinical treatment plan.
[0058] 10) The report is returned and may be downloaded by the
client on a regular schedule, using the packaging and transmission
software for viewing and printing the report by the client at the
recording site. PDF files are opened and displayed using an
interface to Adobe Acrobat Reader (TM) software. Reports may be
printed on any operating system compatible printer.
[0059] 11) Follow up EEG recordings can then be used to track
changes produced by administration of medications by repeating the
entire process outlined above. For follow up studies, the patient
also is interviewed by the treating physician and Clinical Global
Improvement (CGI) is scored. A score of -1 indicates an adverse
effect, 0 no improvement, 1 minimal or mild improvement, 2 moderate
improvement, and 3 marked improvement or remission of symptoms. The
CGI scores are sent to the processing center and are reported along
with changes, expressed as difference scores, on variables shown in
Tables 1-4 above.
[0060] The invention described and claimed herein is not to be
limited in scope by the preferred embodiments herein disclosed,
since these embodiments are intended as illustrations of several
aspects of the invention. Any equivalent embodiments are intended
to be within the scope of this invention. Indeed, various
modifications of the invention in addition to those shown and
described herein will become apparent to those skilled in the art
from the foregoing description. Such modifications are also
intended to fall within the scope of the appended claims.
[0061] The entire disclosures of references cited herein are
incorporated herein, in their entireties, for all purposes.
[0062] Citation or identification of a reference in this
application or in connection with this application shall not be
construed that such reference is available as prior art to the
present invention.
* * * * *