Spatial evolution of neural activity
United States Patent: 7,658,912
Issued: February 9, 2010
Inventors: Paskavitz; James
F. (Holden, MA), Sweet; Larry H. (Providence, RI)
Assignee: University of
Massachusetts (Boston, MA)
Appl. No.: 10/871,745
Filed: June 18, 2004
Training Courses --Pharm/Biotech/etc.
A method for identifying a candidate drug
includes administering a test compound to test subjects and having the
test subjects perform a sustained task. A first evolution of neural
activity in the test subjects is then determined and compared with a
second evolution of neural activity. This second evolution of neural
activity is obtained from control subjects performing the sustained task
in the absence of the test compound. The test compound is then designated
to be a candidate drug when a difference between the first and second
evolutions of neural activity is above a difference threshold.
Description of the
FIELD OF INVENTION
The invention relates to the observation of the spatial evolution of
neural activity, and in particular, to the use of such observations for
development of drugs for treating neurological disorders.
The process of discovering new drugs for treatment of medical disorders
often requires administering hundreds of test compounds to human
volunteers. If the test compound relieves the manifestations of the
disorder, it is considered to be a candidate drug and subjected to further
For many disorders, the manifestations of the disorder are readily
observable. However, in the case of neurological disorders, the
manifestations are more difficult to quantify. For example, if one were to
give a test compound to a volunteer afflicted with Alzheimer's, one would
want to test the effect on that volunteer's short term memory. Each such
test is a time-consuming proposition. Adding to this difficulty is the
need to test enough volunteers to ensure that the conclusions drawn from
these tests merit statistical significance.
Other neurological disorders, for example epilepsy, are characterized by
transient episodes. The search for drugs having therapeutic value in
treatment of epilepsy thus requires long periods of time during which one
must somehow monitor the frequency, duration, and severity of seizures.
The search for drugs effective in suppressing symptoms of schizophrenia
requires somehow determining the extent to which a volunteer's perception
of reality differs from reality, a measure that is fraught with subjective
The invention is based in part, on the recognition that the spatial
evolution of neural activity provides insight into the potential
effectiveness of a test compound at treating a particular neural disorder.
This avoids the difficulties associated with observing gross behavioral
manifestations of the disorder.
In one aspect, the invention includes a method for identifying a candidate
drug. The method includes administering a test compound to test subjects
and having the test subjects perform a sustained task. A first evolution
of neural activity in the test subjects is then determined and compared
with a second evolution of neural activity. This second evolution of
neural activity is obtained from control subjects performing the sustained
task in the absence of the test compound. The test compound is then
designated to be a candidate drug when a difference between the first and
second evolutions of neural activity is above a difference threshold.
In one practice of the invention, the test subjects are selected from
among those afflicted with a neurological disorder for which the candidate
drug is intended to have therapeutic value, and the control subjects are
selected from among those that are free of the disorder, Exemplary
disorders include Alzheimer's disease, Parkinson's disease, and
A variety of sustained tasks can be performed, depending on the type of
neural activity that is to be observed. For example, the sustained task
can be a cognitive task, an N-back task, a semantic reasoning task, or a
visuospatial recognition task.
A variety of imaging methods can be used. For example, in one practice of
the invention, determining a first evolution of neural activity includes
obtaining a sequence of magnetic resonance images of each of the test
subjects during performance of the sustained task.
In another practice of the invention, determining a first evolution of
neural activity includes collecting task data indicative of evolution of
neural activity during performance of the sustained task. The task data is
then filtered to remove contributions arising from background neural
In some practices of the invention, the background neural activity
includes neural activity associated with performance of a reference task.
However, in other practices, the background neural activity includes
neural activity associated with a selected portion of the test subject. In
particular, the background activity can be associated with the selected
portion during performance of the sustained task.
Alternatively, determining a first evolution of neural activity can
include having the subjects perform the sustained task during a first
plurality of task intervals, each having at least a first time segment and
a second time segment. Data from the first and second time segments is
then collected into respective first and second data sets. These data sets
are then filtered to remove a contribution arising from background neural
activity. Such filtering can be performed by, for example, performing a
correlation between data representative of the background neural activity
and the first and second data sets, or performing a t-test between data
representative of the background neural activity and the first and second
In another aspect, the invention includes a method for evaluating an
effect of a physiologic stimulus on treatment of a neurological disorder,
This method includes comparing evolution of neural activity between first
and second groups, with the subjects in the first group having been
administered the stimulus and the subjects in the second group not having
been administered the stimulus. On the basis of a difference between the
evolution of neural activity, an efficacy of the stimulus is determined.
Certain practices include the additional step of selecting the stimulus to
be a pharmacological agent.
Other practices include those in which comparing evolution of neural
activity includes obtaining functional MRI data for subjects in the first
and second groups.
In some practices, comparing evolution includes having subjects from the
first and second groups perform a sustained task.
In other practices, comparing evolution comprises having subjects from the
first and second groups alternate between performing a sustained task and
not performing the sustained task.
The invention also includes systems having an imaging system and a data
processing system configured to carry out the method recited above.
Unless otherwise defined, all technical and scientific terms used herein
have the same meaning as commonly understood by one of ordinary skill in
the art to which this invention belongs. Although methods and materials
similar or equivalent to those described herein can be used in the
practice or testing of the present invention, suitable methods and
materials are described below. All publications, patent applications,
patents, and other references mentioned herein are incorporated by
reference in their entirety. In case of conflict, the present
specification, including definitions, will control. In addition, the
materials, methods, and examples are illustrative only and not intended to
Referring to FIG. 1 (see Original Patent), a new method 10 for identifying
a candidate drug for treatment of a neurological disorder depends in part
on administering a test compound (step 12) to a test group of test
subjects afflicted with the disorder and visualizing the evolution of
neural activity in those subjects (step 14). This evolution of neural
activity is then compared with the corresponding evolution of neural
activity as observed in a control group of healthy subjects who are free
of the disorder (step 16). To the extent that the evolution of neural
activity observed in the test group differs from that observed in the
control group (step 18), the test compound is considered to be a candidate
drug for treatment of that disorder (step 20). If the evolution of neural
activity observed in the control group is similar to that observed in the
test group, the test compound is removed from further consideration as a
candidate drug (step 22).
As shown in FIG. 2 (see Original Patent), observing the evolution of
neural activity (step 14) in a test subject begins with the preparation of
a portion of the subject for acquisition of sequential images (step 24).
Because of the volume of neural activity that takes place within it, the
brain, or a portion thereof, is a natural choice for observation of neural
activity. However, the method disclosed herein is applicable to
observation of neural activity in any type of neural tissue.
Imaging can be accomplished by a variety of image acquisition systems, for
example, MRI ("magnetic resonance imaging") systems, PET ("positron
emission tomography") scanners, quantitative electro-encephalography ("QEEG")
or magneto-encephalography ("QMEG") systems. Thus, preparation for imaging
can include having the subject's head enter the field of an MRI machine,
injecting a radioactive tracer into the subject and appropriately
positioning the subject within a PET scanner, or attaching electrodes
and/or pick-up coils to appropriate areas of the subject's scalp.
Because of the difference between the paramagnetic properties of
oxygenated blood and deoxygenated blood, MRI systems are particularly
useful for image acquisition. The acquisition of multiple MRI images of
the same region separated in time is often referred to as "functional MRI"
Once the subject is prepared, the image acquisition system begins
acquiring a sequence of data sets during a testing period (step 26). As
shown in FIG. 3 (see Original Patent), a testing period 28 is divided into
one or more task intervals 30 separated from each other by reference
intervals 32. Although the task intervals 30 in FIG. 3 are shown as being
the same length as the reference intervals 32, this need not be the case.
Referring again to FIG. 2, during the task intervals 30, the subject is
asked to perform a task (step 34). Depending on the nature of the disorder
and the portion of the brain in which neural activity is to be stimulated,
the task can be a motor task, or any of a variety of cognitive tasks. For
example, if the disorder is one that affects short-term memory, such as
Alzheimer's disease, the task would be one that is expected to exercise
that memory. An example of such a task is the 2-back test in which a
subject is presented with a stream of symbols and asked to determine
whether a current symbol matches the symbol that preceded the preceding
symbol. For disorders of the visuospatial processing system, the test
subject is asked to perform tasks that test the recognition of symbols or
the identification of missing symbols from a set of symbols. For disorders
of the brain's semantic processing system, the test subject is asked to
perform simple reasoning tasks such as recognizing a presented word,
retrieving from memory an association between that word and a category,
and performing a function indicative of recognition of such an
Referring again to FIG. 3 (see Original Patent), each task interval 30 is
divided into a number of time segments 36A-D. During each time segment
36A-D, the image acquisition system collects a task data set 35 TD.sub.ij,
where the index i refers to the task interval 30 and the index j
identifies the time segment 36A-D within the task interval 30. Task data
sets TD.sub.1j, TD.sub.2j, . . . TD.sub.Nj, from corresponding time
segments will later be combined to form one image that shows neural
activity during a selected time interval following initiation of the task.
For example, each task interval 30 will have a first time segment 36A, a
second time segment 36B, and a last time segment 36D.
As shown in FIG. 4 (see Original Patent), the task data sets TD.sub.1j for
first time segments of each task interval 30 will be averaged together,
with the resulting time segment average 37 to be used in constructing a
first image 38 in the sequence of images 40. Task data sets TD.sub.2j for
second time segments of each task interval 30 will likewise be averaged
together, with the resulting time segment average 42 being used in
constructing the next image 44 in the sequence of images 40.
As the number of time segments 36 per task interval 30 increases, the
temporal resolution with which the evolution of neural activity can be
viewed also increases. On the other hand, as the number of time segments
36 increases, each time segment 36 becomes proportionately shorter. Hence
the amount of data that can be gathered during any one time segment 36
decreases. It will therefore be necessary to have more task intervals 30,
and hence a longer test period 28, to maintain the overall quality of the
While the test subject performs the selected task, a great deal of
background neural activity that is not associated with performance of that
task continues to take place. Since it is only the neural activity
associated with performance of the task that is ultimately of interest, it
is desirable to filter out as much non-task related neural activity as
Referring again to FIG. 3, reference data sets 39 (RD.sub.i) acquired
during reference intervals 32 provide a basis for filtering non-task
related neural activity. Since the task is not being performed during the
reference intervals 32, neural activity during the reference interval 32
provides an indication of background neural activity whose statistical
effects on task data sets 35 can later be removed. Because the reference
data set 39 is intended to represent constant background neural activity,
there is no advantage to dividing the reference interval 32 into time
segments and collecting reference data sets 39 in each such time segment.
Consequently, there is generally only one reference data set 39 per
reference interval 32. During the reference interval 32, the subject is
asked to perform a reference task (see FIG. 2, step 41). This process is
repeated, with the subject performing tasks during the task interval (step
34) and performing a reference task (step 41) during the reference
interval 32, until the completion of data acquisition (step 43). The
remaining steps in the new method are to remove the noise from the data
(step 45) and to form images therefrom (step 47).
Referring again to FIG. 4 (see Original Patent), the reference data sets
39 are likewise averaged together. The resulting reference average 46 is
statistically combined with each of the four time segment averages 37, 42,
48, 50 to extract only that data that represents neurological activity
associated with performing the task. A variety of known statistical
techniques are available for achieving this result. For example, one can
perform a cross-correlation or T-test between the time segment averages
and the reference average. Or, one can perform multiple regression
analysis or any one of a variety of non-parametric statistical procedures
to achieve this same goal. In addition, for each image 38, 44, 52, 54 in
the image sequence 40, one might simply evaluate differences between the
reference average 46 and the time segment average 37, 42, 48, 50 for that
image 38, 44, 52, 54.
Whichever statistical technique is used, the end result is to distinguish
a signal due to task-related activity from a reference signal that, in
this case, corresponds to background activity. However, there is no
requirement that the reference signal correspond to background activity.
For example, in some cases, it may be desirable to identify portions of
the brain whose neural activity is correlated with a reference portion of
the brain. In such a case, the reference signal would be a measure of the
time-varying neural activity within the reference portion of the brain.
In the case in which the reference signal corresponds to activity in a
reference portion of the brain, the reference interval 32 is no longer
necessary for collecting reference information. However the reference
interval 32 may still be necessary to allow the test subject to rest,
thereby allowing neural activity to die down so that the task can be
always be repeated against the same backdrop of neural activity.
In this case, in which the image sequence 40 shows neural activity that
corresponds to neural activity in a reference portion of the brain, both
reference data sets 39 and task data sets 35 are collected during
performance of the task, i.e. during the task interval 36. Data arising
from the reference portion of the brain is sequestered from data arising
from the remainder of the brain and processed as described above in
connection with processing the task data sets 35. In effect, the only
distinction is that the reference data set 39 no longer represents the
generally constant background neural activity present when the test
subject is not performing a task. Instead, the reference data set 39 now
represents the time-varying neural activity of a selected portion of the
brain during performance of the task. The distinction is thus analogous to
the difference between determining the trajectory of a moving target
relative to a stationary background and determining the trajectory of a
moving target relative to a moving object.
The procedures set forth above are also applicable to the collection of
data indicative of evolution of neural activity in healthy subjects from
whom the test compound is withheld. Such data, collected in the manner set
forth above, can be used as a basis for comparison to determine whether a
particular test compound is effective.
Observation of the spatial evolution of neural activity as discussed above
has applications other than screening drugs. For example, such
observations may assist in diagnosing a disorder or, in the case of
patients already known to have the disorder, monitoring the disorder. For
patients already being treated for a disorder, such observations can be
used to monitor the effectiveness of the treatment. In addition, the
methods described herein can also be used to identify suitable test
subjects for drug screening studies.
Claim 1 of 11 Claims
1. A method for identifying a candidate
drug for the treatment of Alzheimer's disease, the method comprising:
administering a test compound to test subjects; having the test subjects
perform a sustained N-back task; determining a first evolution of neural
activity in the test subjects; comparing the first evolution of neural
activity with a second evolution of neural activity, wherein the second
evolution of neural activity is obtained from control subjects performing
the sustained task in the absence of the test compound; and designating
the test compound to be a candidate drug when a difference between the
first and second evolutions of neural activity is above a difference
If you want to learn more
about this patent, please go directly to the U.S.
Patent and Trademark Office Web site to access the full