Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
CA 02833398 2013-11-20
METHOD OF IDENTIFYING AN INDIVIDUAL WITH A DISORDER OR EFFICACY OF A
TREATMENT OF A DISORDER
FIELD OF THE INVENTION
This invention relates to the field of neuropsychiatric testing and in
particular to the use of point-
of-gaze data to identify an individual with a psychiatric disorder or
predicting the efficacy of a
treatment of a disorder such as predicting the efficacy of a drug treatment in
depression through
the statistical analysis and modeling of eye movements and point-of-gaze data.
BACKGROUND TO THE INVENTION
The current standard in assessments of neuropsychiatric disorders includes
questionnaires that
require verbal interaction with the person (clinician, caregiver etc.) who is
conducting the
assessment. Examples of commonly used questionnaires for the assessment of
neuropsychiatric disorders are the 21-item Hamilton Depression Rating Scale
(HAM-D) to
assess depression, the 26-item Eating Attitudes Test (EAT-26) to assess eating
disorders and
the Neuropsychiatric Inventory (NPI) to assess behavioural disturbances in
dementia patients.
If subjects minimize or amplify the severity of their symptoms or are unable
to provide accurate
descriptions of their symptoms (e.g., Alzheimer patients) the accuracy of the
assessments is
compromised. Also, due to the time lag between the start of a therapeutic
regime and alleviation
of symptoms patients are often unaware of the effects of the therapy and
cannot provide reliable
information regarding the efficacy of the therapy during the early stages of
treatment. The
current standards of psychiatric assessments may be inaccurate and incomplete.
Recent
research suggested that visual scanning parameters may provide objective
markers that can
support a more accurate assessment of neuropsychiatric disorders and better
prediction of the
efficacy of therapeutic approaches in patients.
Visual scanning devices and parameters derived from the analysis of visual
scanning have been
utilized when viewing images for a variety of purposes. For example, United
States Patent No.
US 7,857,452 relates to a method and apparatus for identifying the covert foci
of attention of a
person when viewing an image or series of images. The method includes the
steps of
presenting the person with an image having a plurality of visual elements,
measuring eye
movements of the subject with respect to those images, and based upon the
measured eye
movements triangulating and determining the level of covert attentional
interest that the person
has in the various visual elements.
8778843.4
CA 02833398 2013-11-20
United States Patent Application Publication No. US 2011/0270123 Al relates to
a method and
apparatus of utilizing an eye detection apparatus in medical application,
which includes
calibrating the eye detection apparatus to a user; performing a predetermined
set of visual and
cognitive steps using the eye detection apparatus; determining a visual
profile of a workflow of
the user; creating a user-specific database to create an automated visual
display protocol of the
workflow; storing eye-tracking commands, for individual user navigation and
computer
interactions; storing context-specific medical application eye-tracking
commands, in a database;
performing the medical application using the eye-tracking commands; and
storing eye-tracking
data and result of an analysis of data from performance of the medical
application, in the
database. The method includes performing an analysis of the database for
determining best
practice guidelines based on clinical outcome measures.
Furthermore, United States Patent No. US 7,046,924 relates to a method for
determining an
area of importance in an archival image. In accordance with this method, eye
information
including eye gaze direction information captured during an image capture
sequence for the
archival image is obtained. An area of importance in the archival image is
determined based
upon the eye information. Area of importance data characterizing the area of
importance is
associated with the archival image.
Moreover, United States Re-issued patent No. US RE39,539 E illustrates an
apparatus for
monitoring movement of a person's eye to monitor drowsiness.
Also, United States Patent No. US 7,206,022 relates to a camera system
provided having an
image capture system adapted to capture an image of a scene during an image
capture
sequence and an eye monitoring system adapted to determine eye information
including a
direction of the gaze of the eye of a user of the camera system. A controller
is adapted to store
the determined eye information including information characterizing eye gaze
direction during
the image capture sequence and to associate the stored eye information with
the scene image.
United States Patent Application Publication No. US 2009/0012419 Al discloses
a system and
method for performing physiological assessments.
United States Patent application Publication No. US 2008/0255949 Al shows a
method and
system for measuring non-verbal and pre-conscious responses to external
stimuli.
United States Patent Application Publication No. US 2007/0066916 Al relates to
a system and
method for determining human emotion by analyzing eye properties.
2
8778843.4
CA 02833398 2013-11-20
United States Patent No. 4,889,422 filed on January 28, 1986 relates to a
diagnostic device and
method for detecting various neurological conditions, particularly dyslexia.
Eye movement
patterns of the subject to be tested are separated into saccadic movements
(both progressive
and regressive), vergence, pursuit movements and fixations, and the subject's
specific eye
movement pattern, as evaluated against a specific stimulus, and normal
patterns is used for
diagnostic purposes. A variety of eye movement detectors is disclosed,
together with a sampling
means which evaluates the eye position at intervals of less than 10
milliseconds.
United States Patent Application Publication No. US 20130090562 Al relates to
methods and
systems for assessing cognitive function. The method includes the steps of
presenting a
plurality of images, wherein the plurality of images comprises a first subset
of images and a
second subset of images; monitoring eye movements of the subject during
presentation of the
first subset of images to obtain first eye movement data; monitoring eye
movements of the
subject during presentation of the second eye movement data; comparing the
first eye
movement data and the second eye movement data to determine an index of
cognitive function;
and correlating the index of cognitive function with a degree of cognitive
function in the subject,
thereby assessing the cognitive function. Wherein the monitoring steps are
carried out using an
optical eye tracking system.
SUMMARY OF THE INVENTION
It is an object of this invention to provide an improved method and system of
identifying
individuals with a neuropsychiatric disorder or to predict and determine the
efficacy of treatment
of the disorder.
It is an aspect of the invention to provide a method of identifying
individuals with
neuropsychiatric disorders or to predict and determine the efficacy of
treatment of the disorder
or detecting individuals who suffer a trauma to the brain by acquiring
information about visual
scanning behaviour and fluctuations of visual scanning behaviour of
individuals comprising
presenting to the individual a sequence of visual stimuli, wherein each visual
stimulus is
comprised of multiple images with specific characteristics, measuring the
point-of-gaze of said
subject on the visual stimuli and calculating a set of statistical measures
that describes the
visual scanning behaviour of the individual on images or portion of images
with the same
characteristics; and making a determination of biases in visual scanning
behaviour of the
individual, by comparing the statistical measures of the individual to the
statistical measures of
controls.
3
8778843.4
CA 02833398 2013-11-20
It is another aspect of the invention to provide a method of identifying the
efficacy of a drug
treatment for an individual with neuropsychiatric disorder comprising
presenting to an individual
undergoing treatment with said therapy a sequence of visual stimuli, wherein
each visual
stimulus is comprised of multiple images with specific characteristics,
measuring the point-of-
gaze of the individual on the visual stimuli and calculating a set of
statistical measures that
describes the visual scanning behaviour of the individual on images or portion
of images with
the same characteristics; and making a determination of changes to visual
scanning behaviour
of the individual by comparing to either the visual scanning behaviour when
not undergoing
treatment with said therapy or the visual scanning behaviour of the individual
during said
therapy; or the visual scanning behaviour of controls.
It is another aspect of the invention to provide a method of screening
individuals for
neuropsychiatric disorders comprising presenting to individuals sequences of
visual stimuli,
wherein each sequence is designed to identify a specific neuropsychiatric
disorder or a specific
symptom that is associated with a neuropsychiatric disorder; measuring the
point-of-gaze of the
individual on the visual stimuli and calculating a set of statistical measures
that describes the
visual scanning behaviour of the individual on image or portions of images
with the same
characteristics, for each of the sequences of visual stimuli; making
determinations of biases in
visual scanning behaviour of the individual for each of the different
sequences of visual stimuli,
by comparing the visual scanning behaviour parameters for each sequence with
those of control
groups; and combining the set of determinations of visual scanning behaviour
biases to provide
an objective quantitative measure of the patient's neuropsychiatric profile.
It is another aspect of the invention to provide a system for identifying an
individual with a
neuropsychiatric disorder or predict and determine the efficacy of treatment
of a disorder
comprising a visual scanning device to collect point-of-gaze data of
individuals, compute
parameters of visual scanning behaviour from the collected data, and compare
the visual
scanning behaviour parameters with those of controls.
It is yet another aspect of the invention to provide a system for identifying
an individual with a
neuropsychiatric disorder or determining the efficacy of treatment of a
disorder comprising: an
eye tracking system to monitor the visual scanning behaviour of individuals,
and a computing
device that includes a presentation module to configure to present visual
stimuli on the monitor
a data analysis module to compute statistical measures of the individual's
visual scanning
parameters from the eye trackers point-of gaze data and to compare these
measures with data
from controls.
4
8778843.4
CA 02833398 2013-11-20
BRIEF DESCRIPTION OF THE FIGURES
Figure 1: Picture of an eye-tracking system and presentation monitor. The
video camera of the
eye-tracking system is in the monitor's stand and the infra red illuminators
are in the two vertical
opaque columns that are attached to the monitor.
Figure 2: Screen caption of the operator's monitor. Estimated gaze positions
are shown in "real-
time" in the upper-right hand corner of the screen. Total time spent on each
of the four images
on the current slide is presented in the bottom-right hand corner of the
screen. The images
from the eye-tracker's camera, metrics of eye-tracking quality, and eye-
tracker controls are
displayed in the bottom-left.
Figures 3(a) and 3(b) is a chart representing that each fixation can be
characterized by a set of
parameters, and each fixation can be linked to a specific image or area of
interest within an
image.
Figure 4 is an example of gaze processing data.
Figure 5 is a diagram of a fixation sequence illustrating an example of data
processing.
Figure 6 is a general block diagram for data processing during the assessment
task.
Figure 7 is an illustration of a slide with images of thin body shapes and
social interactions.
Figure 8a and 8b are histograms of the difference in relative fixation times
on images with thin
body shapes and images with social interaction for patients with Anorexia
Nervosa and controls,
respectively.
Figure 9 shows the differences between the relative fixation times on images
with thin body
shapes and images with social interactions for individuals as a function of
their EAT scores.
Figure 10 is an example of a test slide for the prediction of drug efficacy in
depression.
Figure 11 HAMD scores for patients who responded to the medication
(responders) and for
patients who did not respond to the medication treatment (non-responders). The
scores are
provided for the full length of the study (8 weeks).
Figure 12 illustrates the relative fixation times on dysphoric images of
responders and non
responders prior to and during the drug treatment.
5
8778843.4
CA 02833398 2013-11-20
Figure 13 illustrates the number of visits to dysphoric images versus the
number of fixations
during the first visit to dysphoric images by patients with major depression
disorder.
Figure 14 illustrates the relative fixation times of apathetic Alzheimer
patients, Alzheimer
patients and age matched controls on social, dysphoric and neutral images.
DETAILED DESCRIPTION OF THE INVENTION
In humans, detailed information (high resolution, colour, etc) of objects in
the visual field
(approximately 1800*1350) is obtained by moving the eyes so that images of
these objects fall
on the rod-free, capillary-free portion of the retina - the foveola- (0.3 mm,
approximately 1 ).
During visual exploration, patterns of visual scanning are formed by
successive periods of
steady gaze (fixations) and rapid movements (saccades). Fixations allow areas
in the subject's
visual field to be viewed by the fovea, providing the visual system with high-
acuity color-rich
information, while less detailed information is collected by parafoveal and
peripheral retinal
fields. The patterns of movement (visual scanning behaviour) provide
continuous records of
regions in the visual field that are considered relevant by the subject.
Visual scanning behaviour (VSB) is controlled by both low-level perception
processes (e.g.,
temporal and spatial characteristics of the visual stimuli) and high-level
cognitive processes,
which are driven by the subject's memories, emotions, expectations and goals.
As such, VSB is
affected by many of the processes that interact and contribute to the
development and
maintenance of neuropsychiatric disorders. VSB provides not only behavioural
end products of
cognitive processes but a continuous measure of attention (Herrnans et al,
1999; Toh, Rossell,
and Castle, 2011) that provides clues to the process through which these
products are
achieved. During natural viewing subjects are unaware of their eye movements
and since VSB
can be monitored without requests for meta-cognitive reports or other overt
responses it
provides information that generally can not be observed through the monitoring
of patients'
conscious behaviour.
The present invention uses parameters derived from the subject's VSB when
viewing images to
identify individual with neuropsychiatric disorders or to predict the efficacy
of a treatment of a
disorder. Examples of neuropsychiatric disorders and treatments that are
provided in this
document using the methods and systems of the present invention include, but
not limited to,
treatment of depression with antidepressant medication (SNRI), identifying
patients with
Anorexia Nervosa and treatment of Anorexia Nervosa in a specialized intensive
program at the
hospital and identifying apathy in Alzheimer patients.
6
8778843.4
CA 02833398 2013-11-20
METHODS OF ASSESSING NEUROPSYCHIATRIC DISORDERS
The method includes an eye tracking module configured to monitor the gaze
position of the
subject during viewing of visual stimuli, and a presentation module to present
visual stimuli to
the subject, wherein each visual stimulus is comprised of multiple images with
specific
characteristics. Since patients with neuropsychiatric disorders tend to
selectively attend to
disorder-relevant visual stimuli, often independently of awareness or intent
(Mathews and
McLeod, 1994), if the specific characteristics of the images are relevant to
the disorder being
probed, individuals with such a disorder will exhibit biases in their VSB when
compared with
individuals who do not suffer from the disorder. A computing module linked to
the eye-tracking
module to receive the gaze position data from the eye tracking module is
utilized, to analyze the
data and to derive a set of visual scanning parameters, and to compare the set
of parameters
for an individual with those of control subjects.
Figure 1, shows an embodiment of such a system that was developed to carry out
the
neuropsychiatric assessment. The system (for example Visual Attention Scanning
Technology
(VAST) developed by EL-MAR Inc, Toronto, Ontario, Canada) includes a gaze
estimation
system that records the participant's eye movements. However, other VSB
systems can be
used. The gaze estimation system (for example, EL-MAR Model VISION 2020RB)
uses a digital
video camera and multiple infra-red light sources to illuminate the patient's
face. Images from
the digital cameras are analyzed in real time by algorithms that were
optimized to detect and
estimates eye features [Guestrin and Eizenman, 2006, Guestrin and Eizenman,
2008]. The
estimation of the point-of-gaze is derived from the estimation of the center
of the pupil and
comeal reflections (virtual images of light sources that Illuminate the
subject's face) in the
images [Guestrin and Eizenman, 2006, Guestrin and Eizenman, 2008]. When a
single camera
is used, a calibration routine in which the subject is looking at several
points (3-9) on the
computer monitor has to be completed before the eye-tracker can be used to
estimate gaze
position accurately. When pairs of stereo-images are used a much simpler one-
point calibration
routine has to be completed before the eye tracking system can be used. The
tracking range of
the system described in this embodiment is 30 , and it can accommodate head
movements
within a volume of 27000 ans. This allows participants to move their heads
freely within 1 cubic
foot, which supports natural viewing of the visual stimuli, Visual stimuli are
presented on a
computer monitor (for example, a 19 inch monitor) and each visual stimulus
(slide) includes
several distinct images. A second monitor is used by the VAST system (VAST is
a trade mark of
EL-MAR Inc.) to display the subject's fixation points on each image and
provides the operator
with an interface to control the experimental procedures (Figure 2).
Participants view a
7
8778843.4
CA 02833398 2013-11-20
sequence of slides and the eye movement data is collected by a computer. The
presentation,
recording, and analysis of visual scanning parameters are controlled by
computer programs in
VAST.
The computer program product for assessing neuropsychiatric function comprises
of code,
which when loaded into memory and executed on a processor of a computing
device, is
designed to carry out a method to identify individual with neuropsychiatric
disorders or to predict
the efficacy of a treatment of a disorder comprising the steps of:
a) Presenting images;
b) Monitoring and recording eye movements during the presentation;
c) Analyzing the data to obtain a set of parameters that characterizes the
VSB;
d) Comparing the set of VSB parameters with those of controls.
The method of the present invention can use computing devices that include,
but are not limited
to, desk-top computers, portable computers, mobile computing devices such as
tablets or cell
phones with either internal eye-tracking devices (i.e., eye-tracking devices
that are supported by
the operating system of the computing devices) or eye-tracking devices that
are external to the
computing device (eg., data from the eye-tracker is transferred through one of
the
communication ports of the computing device).
The method of the present invention relies on the collection and analysis of
the subject's visual
scanning patterns when viewing visual stimuli. The collected gaze information
is first divided to
a set of discrete fixations. Fixations can be identified, for example, by
clusters of gaze points
that are within a specific distance (e.g. 1 degree) from each other for a time
period that is
greater than a minimum (eg., 200 milliseconds). Each fixation can be
characterized by a set of
parameters (Fig 3(a)) such as: mean position on the display, duration and the
order in the
sequence of fixations from the time that a visual stimulus was presented. Each
fixation is linked
to a specific image on the display so that the fixation behaviour can be
analysed with respect to
the defining characteristics of images presented to the subject (see Fig.
3(b)) and with respect
to the defining characteristics of specific regions (areas of interest) within
an image. Some of
these characteristics can be used to normalize the fixation behaviour (e.g.,
saliency).
The set of parameters defined in 3(a) is computed as follows: Fixation
position is the average
position of all the eye-position estimates that constitute the fixation.
Fixation standard deviation
is the standard deviation of all the eye-position estimates that constitute
the fixation. Fixation
start and fixation end are the time is milliseconds from the start of the
experiment to the
8
8778843,4
CA 02833398 2013-11-20
beginning and end of each fixation, respectively. The fixation order is the
order of the fixation in
the sequence that started with the first fixation on a new visual stimulus
(eg., slide). The
average pupil-size is the average of pupil-size estimates for data points that
constitute the
fixation.
If the average fixation position falls within the boundaries of an image or an
area of interest
(A01) within an image, the characteristics of the image (e.g., valence,
complexity) and the A01
within the image (e.g., color, corners) are recorded as part of the
description of the fixation.
Example of the processing of gaze position data is described in Fig. 4. The
top row in Fig. 4
shows the order in which discrete fixations are detected. The second row from
the top shows
the images that are linked to each specific fixation and the third row from
the top in Fig. 4 shows
the area of interest in each image that is linked to each specific fixation.
Using this procedure,
visual scanning behaviour can be characterized both in terms of spatial
fixation behaviour (i.e.,
fixations on specific areas of the visual stimulus), temporal fixation
behaviour (i.e., fixations
during sub-intervals of the scanning sequence) and event related fixation
behaviour (i.e.,
fixations following a specific event). For example, as shown in Figure 4, to
determine the fixation
behaviour during the second time that area of interest (a) of Image 1 was
fixated on, one will
look at the characteristics of the Eith fixation. As an example of an event
related fixation
behaviour one can characterise the fixation behaviour following a fixation on
A01 (a) of Image 1.
In the example of Fig, 4, the fixation following a fixation on A01 (a) of
image 1 was always to
A01 (b) of image 1.
Fig. 4: shows processing the gaze position data to create a sequence of
fixations linked to
images or areas of interest within images. Visits are defined by all the
fixations on a specific
image or A01 that occur without leaving the image or the A01 to look at
another image or A01.
A selection of visual scanning behavior parameters are defined as follows.
This list is not
intended to be limiting. These parameters pertain to VSB on whole images or
areas of interest
within images for the total presentation time of each stimulus or sub-
intervals within the total
presentation time (e.g., Image 3 Visit 1).
= Total number of fixations on each image or A01 within an image during
each slide
presentation.
= Relative number of fixations: Total number of fixations on each image or A01
within an
image divided by the total number of fixations on all images or AOls on the
slide.
9
8778843.4
CA 02833398 2013-11-20
= Total duration of fixations on each image or AOls within images during
each slide
presentation.
= Relative duration of fixations: Total duration of fixations on each image
or A01 within an
image divided by the total number of fixations on all images or AOls on the
slide.
= Total number of visits: A visit to an image or an A01 on an image starts
with a fixation within
the boundaries of this image or the A01 within the image and ends with a
fixation outside the
boundaries of this image or the A01 within the image (see Figure 5). The total
number of
visits to an image or an A01 within an image is the number of visits during
the whole slide
presentation.
= Relative number of visits: Total number of visit to each image or an A01
within art image
divided by the number of visits to all images or AOls on the slide.
= Number of fixations within each visit.
= Glance duration: The length of time of all fixations within each visit.
= Glance duration before all images on the slide were seen.
= Average glance duration: defined as the sum of all glance durations during
each slide
presentation divided by the number of visits.
= Average glance duration before all images on the slide are seen.
= Average glance duration after all images on the slide are seen.
= Temporal fixation order.
= Transition probabilities from an image or an A01 within an image to another
image or
another A01.
= Total scan path within an image or an A01 within an image. The sum of all
horizontal and
vertical eye movements within an image or an A01 within an image during the
presentation
of a slide.
= Scan path within visit. The sum of all horizontal and vertical eye movements
within an image
or an A01 within an image during each visit.
= Total dispersion of scan path: The variance of fixation positions within
an image or an A01
within an image during the presentation of a slide.
= Dispersion within a visit; The variance of fixation positions within an
image or an A01 within
an image during each visit.
8778843,4
CA 02833398 2013-11-20
= Average dispersion per visit: The sum of all dispersions within visits
divided by the number
of visits.
Figure 6 provides a general block diagram for the processing sequence of a
neuropsychiatric
assessment. During the course of the assessment task, a series (N in Figure 6)
of visual stimuli
(e.g., slides) is presented to the subject. The number of visual stimuli and
the duration that each
visual stimulus is determined by the assessment being conducted. For example,
50 slides are
presented for 10.5 seconds each, for the assessment of the efficacy of
treatment in depression
while for the assessment of eating disorders 37 slides are presented for 12
seconds each.
In accordance with the present invention images on each slide are classified
according to
characteristics that include, for example, valance, arousal, dominance,
complexity and thematic
content (eg., sad, happy, angry, violence, suicide, neutral, thin, fat etc.).
In accordance with the present invention for each visual stimulus (eg., slide)
gaze position data
is processed to generate a set of VSB parameters for each image and/or A01
within an image,
for the whole presentation interval or any sub-interval of the presentation
and for all defined
event related fixations. Following the processing of data from all visual
stimuli (eg., slides),
statistical descriptions (mean, median, Standard deviation) of VSB parameters
for all images
that share the same characteristics are computed. For each individual, the
statistical description
for one type of images (eg., sad, happy) can be normalized ,for example, by
subtracting the
same statistical description for another type of images (eg., neutral). For
example, if individuals
tend to have short fixations or long fixation they will exhibit these patterns
of fixations on both
sad images and neutral images. By subtracting the VSB parameters of neutral
images from that
of sad images the manner in which individuals tend to scan an image, which
might be
independent of the content of the images, will be minimized.
In accordance with the present invention for each visual scanning parameter or
a set of visual
scanning parameters statistical tests that compare the statistical
description(s) of the VSB of the
individual being tested with the statistical description of the VSB of control
groups to determine
if the individual suffers from the specific disorder that the assessment task
is designed to
identify. The control group can be a group of individuals that do not suffer
from the disorder that
the assessment task is designed to identify or/and a control group of
individuals that suffer from
the disorder that the assessment task is designed to identify. In one
embodiment of the
assessment task, if the visual scanning parameter being tested falls outside
the range defined
by the mean of a control group r a (where r is a constant to be decided for
each assessment
task and a is the standard deviation of the value of the parameter for control
subjects) and
11
8778843.4
CA 02833398 2013-11-20
inside the range defined by the mean r a of a control group of individuals
that suffer from the
disorder, the assessment task indicates that the individual suffers from the
neuropsychiatric
disorder that the assessment test is designed to identify.
In another embodiment of the present invention for each visual scanning
parameter or a set of
visual parameters, statistical tests compare the statistical descriptions of
visual scanning
parameters for the subject being tested before the start of treatment, for
example, with the
statistical description of the same subject during or after treatment for
example.
The methods of the present invention can be carried out to assess many
neuropsychiatric
functions. Non-limiting examples of such assessments are set out below.
IDENTIFYING INDIVIDUALS WITH ANOREXIA NERVOSA AND DETERMINING THE
EFFICACY OF TREATMENT IN PATIENTS WITH ANOREXIA NERVOSA
Introduction
Anorexia Nervosa (AN) is a severe and chronic neuropsychiatric disorder with
one of the
highest mortality rates of any psychiatric illness. AN is characterized by
food restriction leading
to weight loss, an extreme fear of fat or weight gain despite being
underweight, body image
distortion and a denial of the severity of the illness (American Psychiatric
Association, 2000).
Self-reported clinical assessment instruments, such as the Eating Attitudes
Test (EAT-26)
(Garner et. al., 1982), that are often used as screening tools are limited in
utility when
participants minimize or misrepresent their behaviour. It is estimated that
only 30% of the
population of people with AN actually receive treatment (Preti, et al., 2009)
and the lack of
objective indicators for anorexia nervosa impact identification, diagnosis and
the course of
treatment of the disorder (Pinhas and Bondy, 2010) .
Traditional cognitive tests to measure attentional biases, such as the Stroop
color naming
interference test (Long et al, 1994) and the Dot-probe test (Shafran et al.,
2007)) use reaction
times to measure attentional biases. These indirect methods provide only a
snapshot of the
processes by which subjects allocate attention (i.e., attention is only
measured at one specific
instant of time). Therefore, these traditional methods suffer from low
sensitivity and specificity. A
more direct method to estimate attentional biases is to measure visual
scanning behaviour
(VSB), which provides continuous record of the attention allocation processes
(Jansen et. al.,
2005, George et. al., 2011, Giel et. al., 2011a, Wietersheim et al., 2012).
Using visual scanning
measures, George et al*, (2011) showed that control subjects fixate mainly on
the subject's
abdominal region while patients with AN have wider fixation pattern that
encompasses another
body features such as hip and collar bones. Jansen et al, (2005) showed that
participants who
12
8778843,4
CA 02833398 2013-11-20
were extremely dissatisfied with their weight or shape focused on the ugly
parts of themselves
and the most attractive parts of others while participants who were satisfied
with their bodies
concentrated on the self-identified ugly parts of others. In a subsequent
study of patients with
AN, Wietersheim et al. (2012) showed that these attentional biases are often
small and vary
greatly. Even though group attentional biases in AN were identified in all of
the above studies,
the large variability between patients and the overlap between the visual
scanning behaviour of
patients and controls did not allow for robust detection (high sensitivity and
specificity) of visual
scanning biases in individual subjects with AN.
Using the methodology described in this invention we were able to show that
patients with AN
have visual scanning behaviour that is significantly different from that of
control subjects. In the
test that we developed to identify patients with AN, participants view visual
stimuli (slides) with
images of thin body shapes alongside images of social interactions for
relatively long time-
periods (12 seconds). Analysis of the measured visual scanning patterns showed
that when
visual scanning behaviour (VSB) of AN patients is compared to the VSB of
control subjects, AN
patients had significantly higher relative fixation times (RFTs) on images
with thin body shapes
and significantly lower RFTs on images with social interactions. The
difference between the
RFTs of AN patients and control subjects is maximized when the RFTs on social
images are
subtracted from the RFTs on images with thin or fat body shapes. We use the
differences
between the RFTs on thin body shapes and images with social interactions in
individual
subjects and a log-likelihood ratio (LLR) processor to detect biases in VSB of
AN patients.
Participants
The patients in the participant behaviours all patients in a specialized
eating disorder program.
They all had confirmed/witnessed behaviours clinically that were consistent
with AN and did not
suffer from depression or OCD co-morbidity. The control group consisted of
individuals who
never had any known eating disorder or any other mental illness and who scored
below the
clinical cut point of 20 on the EAT-26 (Garner, et al., 1982),
A total of 20 patients with a diagnosis of AN and 23 controls participated in
the study. Fourteen
of the 20 AN patients were hospitalized at the time of the test while 6 had
completed treatment
and were in the process of recovery. The mean age of controls (14.4+1,82
years) was not
significantly different from patients (AN: 15.00+1.73 years). The AN patients
who were
hospitalized at the time of the test (AN-pts) were analyzed subsequently
separately from those
that completed the treatment (AN-rec) as they no longer met full criteria for
AN. The mean EAT-
26 score for the AN-pts and control groups were 38.0 22.3 and 6.5 5.6
respectively (t(31) =
13
8778843.4
CA 02833398 2013-11-20
6.07, P = 5.01x104). Four of the AN-pt had a sub-clinical EAT-26 score (<20),
These patients
had either intentionally or unintentionally self-reported behaviour that was
inconsistent with the
other clinical data available and are considered unreliable historians (UH)
for the purpose of the
study. The proportion of UHs within the patient group (30.8%) is similar to
the 27.6% found by
Pryor, Johnson, Wiederman, and Boswell (1995) who described similar finding
that they referred
to as denial.
Visual stimuli
Visual stimuli were organized into slides that were presented on a computer
monitor. Each slide
had four images that were arranged in a 2 x 2 configuration (Figure 7). Slides
had two images of
thin body shapes (e.g., visible rib cage or hipbone, full image of thin
subjects) and two image of
social interactions,
Subjects looked at 37 slides that included 16 test slides and 21 filler
slides. Filler slides intended
to mask the purpose of the experiment and had images that were neutral in
content. The spatial
position of images from each image type (e.g., thin body shapes, social
interactions) were inter-
mixed (i.e., for the set of 16 slides, each category of stimuli appeared in
each quadrant of the
slide the same number of times).
Methodology
The slides were presented on a 19" computer monitor that is part of EL-MAR'S
Visual Attention
Scanning Technology (VAST, EL-MAR Inc. Toronto, Ontario, Canada). VAST
incorporates a
binocular gaze estimation system a real-time processor to estimate a set of
visual scanning
parameters (Eizenman et al., 2003, Hannula et al. 2010) and a monitoring
station to supervise
the procedure. Processing of eye¨gaze data included the segmentation of gaze-
position data to
fixations, the association of fixations with images and the estimation of
visual scanning
parameters. The relative fixation time (RFT) on each image on each slide was
calculated by
dividing the total time of all fixations on the image by the total time of all
fixations on all the
images on the slide. The difference in relative fixation times (RFT) on each
slide was
calculated by subtracting the average RFT on the social images on the slide
from the average
RFT on the images of thin body shapes on the slide. To detect VSB biases, the
ARFTs for all
test slides were processed by a log-likelihood ratio (LLR) processor. The
processor first
determines the likelihood that a measurement of a single ARFT is from an AN
patient or from a
control subject (Equation 1). Then the processor calculates the log likelihood
ratio for the set of
measurements from each subject (equation 2). When the output of the LLR
processor was
14
8778843,4
CA 02833398 2013-11-20
greater than a threshold (log 1 = 0), the processor detected VSB biases that
are consistent with
the VSB of patients with AN.
In accordance with the present invention the LLR processor is one example of a
statistical
procedure to identify an individual with a neuropsychiatric disorder or to
determine the efficacy
of a treatment for the disorder.
The LLR Processor
To determine the likelihood that a measurement of a difference in the relative
fixation time
ARFTi [ARM! )) on slide j, is from a patient with AN is calculated by
the likelihood ratio
):
P(ARFTAClass = AN)
Arj R FT i)1 = __________ j= 1...N (1)
PORFTACLass = Control)
where, j = 1...N are the test slides, P(LRFTiklass = AN)and TilClass =
Control) are the
ORF
conditional probability densities of ARFri for patients with AN and for
controls, respectively.
When A11 is greater than 1, the measurement of ARFri is more likely to come
from a patient
with AN than from a control subject. As visual scanning behaviour (VSB) is
independent from
slide to slide the log-likelihood ratio (LLR) processor for a set of N
measurements is:
log (0/17i ) (2)
When the output of the LLR processor is greater than a threshold (log 1 = 0),
the processor
detects VSB that is more consistent with that of a patients with AN than that
of control subjects.
Figure 8 shows histograms of ARFT for patients with AN (8,) and control
subjects (8b). (i.e.,
slides with thin body shapes and social interactions). The histograms in
Figure 8 were used in
the calculations of the log-likelihood ratios (Equation 1 and 2). For each
measurement, the
conditional probability density was approximated by the height of the bin in
the histogram that
included that measurement, For each subject/patient the histogram were re-
calculated from a
data set that excluded the data for this patient/control, so that the
calculations of the log-
likelihood ratios for each patient/control were not biased by her own data
(leave one out
procedure).
8778843,4
CA 02833398 2013-11-20
Results
Figure 9 shows the differences between the RFTs on images with thin body
shapes and the
RFTs on images with social interactions (ARFTrkin) for individual subjects, as
a function of their
EAT-26 scores. The dashed vertical line at an EAT-26 score of 20 indicates the
clinical cut
point. As a group, ARFroan for AN-pts (M = 0.200, SD = 0.098, range 0.036 to
0.348) is
significantly larger (f (35) = 7.335, P < 0.001) than that of the control
subjects (M = -0.044, SD =
0.099, range -0.226 to 0.140). Also, as a group, the mean ARFTthen of the six
patients who
completed their treatment (M = -0.006, SD = 0.083, range -0.132 to 0.108) is
not significantly
different from the control group (t (27) = -0.878, P = 0.388). The LLR
processor detected biases
in VSB that are consistent with the VSB of AN patients (i.e., the output of
the processor is
greater than 0) in 13/14 (sensitivity 93%) of the hospitalized patients and in
3 of the 23 control
subjects (specificity 87%). Note that a very simple detector (e.g., ARFTthin >
0) can be used to
detect biases in VSB of AN-pts with relatively high sensitivity (100%) and
specificity (65.3%).
The LLR processor optimizes the performance of the detector when the
optimization criteria is
to maximize the sensitivity and specificity of the detector, simultaneously
(i.eõ the sum of
sensitivity and specificity is maximized). VSB biases were detected in three
of the four
hospitalized patients that minimized their symptoms on the EAT-26 test and in
one of the six
AN-rec patients.
Summary
The visual scanning behaviour of AN-pts who were hospitalized during the time
of the tests is
distinctly different from controls and patients that completed their treatment
and are in the
process of recovery. The relative fixation times (RFTs) on images with thin
body shapes of
these patients were higher than those of control subjects or patients that
completed their
treatment while their RFTs on images with social interactions were lower.
Individual AN-pts with
high RFTs on images with thin body shapes and low RFTs on social images had
been ill and in
treatment for over a year, were resistant to recovery clinically or were low
weight at the time of
the study. Clinically, patients that were most rigidly adherent to the
cognitive patterns
(preoccupation with thoughts about shape and weight, self-absorption and
isolation from
external influences) typical to AN had the largest biases in VSB. AN-pts with
lower RFTs on
images with thin body shapes and higher RFTs on images with social
interactions began to
normalize their eating patterns in treatment more readily. As these
characteristics are not
always predictable from the EAT-26 scores, the analysis of biases in VSB is
useful in the
prognosis and treatment of patients with AN.
16
8778843 4
CA 02833398 2013-11-20
Using a LLR-processor, biases in VSB that are consistent with the VSB of AN-
pts were detected
with high sensitivity (93%) when AN-pts looked at slides with images of thin
body shapes and
social interactions. Such biases were detected in only 13% of control subjects
who looked at the
same slides (high specificity - 87%).
Biases in VSB that are consistent with the VSB of AN-pts were detected in
three of the four AN-
pts that either minimized or misrepresented their behaviour (patients with EAT-
26 scores that
are less than 20). The one patient for whom VSB biases could not be detected
was a patient
who was identified very early in her illness (duration of symptoms two
months), had never had a
severely low weight and clinically had relatively mild cognitive symptoms of
weight and shape
concerns. As biases in VSB are a reflection of attentional biases that tend to
occur early in the
information-processing sequence and are often independent of awareness or
intent (Mathews
& MacLeod, 1994) they often bypass the volitional component of self-report
measures and are
less available to conscious manipulation. As such, biases in VSB can provide a
method to
screen adolescents who may be minimizing or misrepresenting the presence of AN
cognitions
and behaviours.
In the six patients who completed an intensive treatment program (AN-rec),
biases in VSB that
were consistent with biases of AN-pts were detected in only one patient. This
patient, whose
score on the EAT-26 test was less than the clinical threshold, completed the
treatment program
only two days before she was tested after being ill for 5 years. AN-rec
patients for whom no
biases in VSB were detected completed the treatment program at least six
months before they
were tested. It is possible that the results for this patient demonstrate a
transitional phase,
where the fixed visual scanning pattern of AN-pts whose main focus is images
with body/weight
shapes is still present, but the patient is already aware of changes induced
by the treatment
program. One of the AN-rec patients for whom no biases in VSB were detected,
had an EAT-26
score that was higher than the clinical cutoff (the EAT-26 score was 41). The
lack of biases in
VSB for this patient might also indicate a transitional phase, where the
visual scanning pattern
became more "normative" before the patients has become completely aware of all
of the
changes in her behaviour.
Observations
Using the methodology of this invention which enhances the differences between
the visual
scanning behavior of AN-pts and control subjects and a standard log-likelihood-
ratio processor,
biases in VSB of individual AN-pts can be detected with high sensitivity (93%)
and specificity
(87%). The inability to detect VSB biases in 83% of the patients who completed
an intensive
17
8778843.4
CA 02833398 2013-11-20
treatment program and were not hospitalized during the time of the study
suggest that VSB
biases are not traits of individual patients but rather reflect the state of
patients during the time
of the test (i.e., the biases can be used to monitor the progress of
treatment), Biases in VSB
were detected in three of the four AN-pts that either minimized or
misrepresented their
behaviour (patients with EAT-26 scores that are less than 20). Since biases in
VSB often
bypass the volitional component of self-report measures and are less available
to conscious
manipulation they might help to identify subjects who are at risk of
developing anorexia nervosa.
Such an objective physiological indicator is important since a significant
proportion of the
adolescent AN population is unable or unwilling to self identify.
Since patients with eating disorders are preoccupied with the ability to
control their eating
habits, using the methodology of this invention with images of low or high
calorie food, for
example, can be used to identify biases in visual scanning behaviour that are
associated with
this pre occupation.. For example, visual stimuli can include images of
positive eating habits
(low calorie food eaten in controlled circumstances) and negative eating
habits (high calorie
foods being eaten in an uncontrolled fashion, e.g." binge like with fingers)
to differentiate
patients with Anorexia Nervosa (AN) from patients with Bulimia Nervosa. While
both groups of
patients tend to avoid food images (when compared to controls), BN patients
tolerate images of
negative eating habits better than patients with AN and therefore have longer
fixation times on
such images (as compared to patients with AN). Using images of low and high
calorie food
provide another objective marker to identify patients with eating disorders
(i.e., a marker that is
different from that obtained by using body shapes) and can be used to
differentiate between
different groups of patients with eating disorder (eg., AN and BN).
By using the method of this invention with different sets of disorder-relevant
visual stimuli (single
or multiple slide presentations), biases in visual scanning behaviour to
visual stimuli (images)
that probe different characteristics of the same disorder (eg., for eating
disorders - body
shapes and low and high calorie food) or different neuropsychiatric disorders
can be obtained.
For example, if the image set that is used to detect apathy (later in this
document) is used with
the image set that is used to identify AN patients, the extent to which
patients with AN also
suffer from apathy can be quantified. By using several sets of images, where
each set probes
a different neuropsychiatric disorder (e.gõ depression, apathy, etc. ), biases
in visual scanning
behaviour that are associated with each neuropsychiatric disorder can be
obtained. The set of
biases in visual scanning behaviour can then provide an objective measure of
the patient's
psychiatric profile (i.e, a model of the patient's psychiatric state).
18
8778843A
CA 02833398 2013-11-20
DETERMINING THE EFFICACY OF DRUG TREATMENT IN PATIENTS WITH MAJOR
DEPRESSION DISORDER
Introduction
Depression is a syndrome that tends to be chronically recurring and affects
about 20% of the
population worldwide. Antidepressant medication, which increases the levels of
certain brain
neurotransmitters (e.g., norepinephrine or serotonin) that are lacking in
depressed people, is the
most common treatment for depression. The most popular class of antidepressant
medication is
SSR1's (selective serotonin reuptake inhibitors - Prozac, Zoloft, Paxil,
Luvox) but there are
several major classes of antidepressant drugs (SNRIs - effexor, Serzone;
Bupropion
Wellbutrin; Mirtazapine ¨ Remeron; TCAs (Tricyclics) - Elavil, Pamelor,
Norpamin; and MAOls
(MAO inhibitors) Pamate, Nardi!). One of the more enduring and problematic
problems in
treating depression is associated with the fact that only 50% of patients
respond to a specific
drug and that the actual therapeutic effect of significant alleviation of
depressive symptoms may
not appear until after 2-6 weeks of daily dosing. The existence of this time
lag and the inability to
predict if drug treatment will be effective for specific patients pose
significant clinical problems. If
the treatment turns out not to be effective, precious treatment time has been
lost, translating into
increased risk for serious consequences and increased suffering for the
patient. A method to
predict whether a treatment was destined to be efficacious and therefore
should be continued,
or if not, should be abandoned in favour of a different drug or drug class
would be of great
value. The method of this invention describes such a method.
Using analysis of natural visual scanning behaviour (i.e., patients look
naturally at images on a
computer monitor), Eizenman et. al. (2003), and later Kellough et al., (2008)
found that when
compared with controls, depressed subjects have longer fixation times on
dysphoric images and
had difficulty shifting attention away from these images. Based on these
initial observations the
methods of this invention were developed to predict the efficacy of treatment
in individuals with
major depression disorder..
Participants
Thirty four patients with Major Depressive Disorder participated in the study.
All patients were
evaluated by a psychiatrist and met the DSM-IV-TR criteria for Major
Depression. Following a
baseline visit (Visit 2), patients received 60 mg Duloxetine antidepressant
monotherapy (SNRI)
PO once daily for the duration of the study.
19
8778843,4
CA 02833398 2013-11-20
Testing Procedure
, _
VISIT V1
(Screen)* V2 V3 V4 V5 V6 V7 V8
Week -1 0 1 2 3 4 5 6
Informed Consent X
Medical & Psychiatric x
Hx
Demographics X
Entry Criteria X
Physical Exam X
Vitals (Wt, Ht) X
Urine Drug Screen X
MINI X
Blood pressure, pulse X X X X
HDRS-17 X X X X X X X
Visual Scanning X X X X X X X
Adverse Events X X X X X X X
Medication Compliance X X X X X X X
µ._
Review of Concurrent
X X X X X X X
Medication
Table 1: Study Flowchart.
Prior to the start of the drug medication treatment, the visual scanning
patterns of each patient
were measured a(V2 visit) Then, each patient had their visual scanning
patterns recorded
once per week for a further six weeks.
Visual Scanning task
Subjects' visual scanning patterns were recorded as they view a presentation
of visual stimuli.
The visual stimuli were organized as a series of slides with four images per
slide. Images were
selected from libraries such as the International Affective Picture System
(lAPS) [Lang et al.,
(1999)] and photos.com. Each image was classified with respect to its thematic
content (e.g,
social interaction, homelessness), complexity (simple to complex, 1-10) and
was rated for
valence and arousal. Valence is a measure of a subject's relative pleasure in
viewing an image,
while arousal classifies subject reaction to an image in the continuum from
relaxed to excited.
Valence, arousal, thematic content and complexity were the criteria used to
select images on
8778643A
CA 02833398 2013-11-20
each slide. The four images for the early detection of drug efficacy in
depression included two
dysphoric images and two images with social interactions. The positions of the
images on the
slide were assigned randomly. Dysphoric images displayed themes of loss and
sadness, illness
and despair while images with social interactions presented themes of
interpersonal attachment
and social content, Dysphoric images were selected to have valence ratings
below 4 (low
valence), while images with social interactions were rated above 6 (high
valence). Images on
the same slide have similar arousals and complexities. An example of a test
slide is shown in
Figure 10.
Each slide presentation included also filler slides that have images with
similar characteristics
(themes, valence, arousal, complexity) and are used in the analysis to
normalize the scanning
patterns of individual subjects. The slide presentation used 15 test slides
with dysphoric and
social themes and 20 filler slides. The positions of the four images on each
test slide were
randomly changed between sessions. Each slide was presented for 10.5 seconds
for a total
presentation time of 8 minutes and 45 seconds. Participants sat at a distance
of approximately
65 centimeters from the monitor so that the visual angle subtended by each of
the four images
on each slide is approximately (15.2 x 11.4 ). The horizontal and vertical
separation between
any two images is greater than 2.5 .
To predict the efficacy of drug treatment we used the following VSB
parameters: 1) the average
number of visits to dysphoric images prior to the start of the medication
treatment 2) the
average number of fixations during visit 1 prior to the start of the
medication treatment, and 3)
the direction of change in relative fixation time on dysphoric images during
the first week of
medication.
Results
After 8 weeks, a psychiatrist classified 17 of the 34 patients who
participated in the study as
responders to the medication (had a HDRS-17 of half their initial value in
their last visit) and 17
patients as non-responders. Figure 11 shows the HORS-17 scores for the two
groups
Figure 11: Means and standard deviations of HDRS-17 scores for responders
(blue) and non-
responders (red) for the 34 patients who participated in the study. After 8
weeks, 17 patients
were classified as responders and 17 patients were classified as non-
responders. Asterisks
indicate significant differences at a level of a=0.05.
21
87788434
CA 02833398 2013-11-20
Figure 12 shows that as a group prior to the start of medication (i.e, before
Visit 3) there are
significant differences (a = 0,05) in the number of visits to dysphoric images
between
responders and non responders. As a group, the number of visits to dysphoric
images of
responders in Week 2 (M = 1.91, SD = 0.94) is significantly higher (a = 0.013)
than the number
of visits to dysphoric images of non responders (M = 1.57, SD = 0.83).
As a group, the number of fixations during the first visit of responders in
Week 2 (M =2.56, SD =
0.13) is significantly lower (a = 0.03) than the number of fixations during
the first visit of non-
responders (M= 3.15, SD=0.209). Using the above two VSB parameters, and a
linear line (see
Figure 12) that defines the boundary of a classifier for responders and non-
responders (patients
with VSB parameters that are above the line are classified as responders) the
test can predict
the response to drug treatment with a sensitivity of 64.7% and a specificity
of 76.5%, prior to the
start of treatment. The classifier has a positive predictive value (PPV) of
73.3% and a negative
predictive value of 68.4 %.
When the direction of changes in the relative fixation times on dysphoric
images during the first
week of treatment (i.e,, if the relative fixation time on dyspohric images is
lower in the third visit
than the relative fixation time in the second visit), is used by the
classifier, the specificity of
classifier is increased to 86.4% and the sensitivity to 58,3% with a PPV of
70% and NPV of
79.2%. These results show that after one week of treatment the
detector/classifier can predict
who will not respond to the drug treatment with high accuracy (86.4%).
IDENTIFYING APATHY IN ALZHEIMER PATIENTS
Introduction
In patients with Alzheimer disease (AD) it is important to differentiate
between apathetic and
non-apathetic patients as the diagnosis of apathy affects the course of
treatment. Assessing
Alzheimer patients is a difficult and imprecise task that requires highly
trained personnel. An
objective method to identify apathy in Alzheimer patients will have
significant clinical
implications for diagnosis and pharmacological treatment, The current
invention provides an
objective method to identify patients with apathy.
Participants
31 patients (ages 77.2 9.2 years) with AD were tested. 15 of the AD patients
had apathy (NPI
¨ apathy > 4 (Robert et al., 2009) and 16 were non-apathetic Alzheimer
patients. Additionally 21
Age matched non-AD participants (ages 71.4 8.4 years) were tested as a
control group. AD
patients were screened for apathy using the Neuropsychiatric Inventory (NPI)
apathy subscale.
22
87788414
CA 02833398 2013-11-20
The Modified Mini Screen (MMS) was administered to healthy controls to exclude
those with
marked neuropsychiatric disturbances. As a group the mean NI31-apathy for the
non-apathetic
AD patients was 0,8 1.1 and for the AD patients with apathy, 5.9 2.6.
Methodology
Participants were tested with the VAST system (EL-Mar Inc. Toronto, Ontario).
The study slides
included 16 test slides and 10 filler slides, The slides were presented
sequentially, with each
slide presented for 10.5-seconds. The total duration of the assessment was
less than 5 minutes.
Each test slide included four images: two images of neutral objects (low
arousal and moderate
valence), one image of social interactions (high valence and high arousal) and
one dysphoric
image (low valence and high arousal). Images were selected from the
standardized
International Affective Picture System (lAPS) (Lang et, al, 1999) and from
photos.com. AD
patients with apathy scan images with different characteristics in a similar
manner so that
differences in VSB parameters when images with large differences in valence
and/or arousal
are viewed, are smaller than the differences observed when non-apathetic AD
patients and age
matched controls view the same images. In this example of the current
invention, the difference
in relative fixation times on images with social interactions and neutral
images is used to identify
AD patients with apathy.
Results
Figure 14, shows the relative fixation times of the three groups of
participants (AD, AD-apathetic
and age-matched controls) on three different types of images (social,
dysphoric and neutral).
Age matched controls and non-apathetic AD patients have similar relative
fixation times on all
three types of images. Apathetic AD patients show significantly different
visual scanning
patterns. Alzheimer patients with apathy have significantly lower relative
fixation times on
images with social interactions when compared with AD patients who are not
apathetic (a =
0.04) and age matched controls (a 0.01) and significantly higher relative
fixation times on
neutral images when compared with age matched controls (a = 0.02).
When differences between the relative fixation times on social images and
neutral images are
used for the identification of apathy (using a naive Baysian Classifier), 67%
of the AD patients
with apathy were identified as apathetic, 72% of the non-apathetic AD patients
were classified
as non-apathetic and 78% of the control subjects were classified as non
apathetic.
23
8778843.4
CA 02833398 2013-11-20
Other
The foregoing embodiments of the invention are examples and can be varied in
many ways.
Moreover, the invention includes detecting individuals who suffer a trauma to
the brain (for
example a concussion, whether from sports or otherwise) and might develop
symptoms similar
to those observed in individuals with neuropsychiatric disorders. Such
variations or
modifications are intended within the scope of the following claims.
24
8778843.4
CA 02833398 2013-11-20
References:
1) American Psychiatric Association (2000). Diagnostic and statistical manual
of mental
disorders: Fourth Edition. Text Revision (DSM-IV). Washington, DC: American
Psychiatric
Association.
2) Bauser, D.A.S., Suchen, B., Duan I., (2011). Differences between perception
of human
faces and body shapes: Evidence from the composite illusion. Vision Research,
51:195-
202.
3) Dobson, KS., Dozois, D.J., (2004). Attentional biases in eating disorders:
a meta-analytic
review of Stroop performance. Clinical Psychology Review, 23(8):1001-1022.
4) Downing, P.E., Jiamg, Y.,rn Shuman, M., Kanwisher, N., (2001). A cortical
area selective
for visual processing of the human body. Science, 293:2470-2473.
5) Eizenman, M., Yu, LH., Grupp, L., Eizenman, E., Ellenbogen, M., Gemar M.,
Levitan,
R.D., (2003). A naturalistic visual scanning approach to assess selective
attention in major
depressive disorder. Psychiatry Research, 118(2):117-128.
6) Fairburn C.G., Cooper L, Shafran R. (2003). Cognitive Behaviour therapy for
eating
disorders: a "transdiagnostic" theory and treatment, Behaviour Research and
Therapy,
41, 509-528.
7) Garner, D.M,, Olmsted, M.P., Bohr, Y., Garfinkel, P.E..(1982). The eating
attitudes test
psychometric features and clinical correlates. Psychological Medicine,
12(4):871-8.
8) Garner, D.M., Vitousek, K.M., Pike, K.M. (1997), Cognitive-behavioral
therapy for anorexia
nervosa. In D.M. Garner & RE. Garfinkel (Eds.), Handbook of treatment for
eating
disorders (pp. 94-144). New York: Guilford Press.
9) Garner, M., Mogg, K., & Bradley, B. P. (2006). Orienting and maintenance of
gaze to facial'
expressions in social anxiety. J Abnorm Psychol, /15(4), 760-770.
10) George H.R., Cornelissen P.J., Hancock P.J.B., Viviniemi V.V., and Tovee
M.J., (2011).
Differences in eye-movement patterns between anorexic and control observers
when
judging body size and attractiveness. British Journal of Psychology, 102, 340-
354.
11) Giel, K.H., Friederich, H.C., Teufel, M., Hautzinger, M., Enck, P. and
Zipfel, S.,(2011a).
Attentional processing of food pictures in Individuals with Anorexia Nervosa ¨
An Eye-
Tracking study. Biological Psychiatry 69, 661-667.
12) Giel, K.H., Teufel, M., Friederich,
Hautzinger, M., Enck, P. and Zipfel, S,,(2011b).
Processing of pictorial food stimuli in patients with eating disorders ¨ a
systematic review.
International Journal of Eating Disorders 44:2 105-117.
28
8778843.4
CA 02833398 2013-11-20
13) Guestrin, ED., Eizenman, M., (2006). General theory of remote gaze
estimation using the
pupil center and corneal reflections. IEEE Transactions on Biomedical
Engineering
53(6):1124-33.
14) E.D. Guestrin and M. Eizenman," Remote point-of-gaze estimation for
studies of selective
attention and mood disorders", Presented at the 30`h Canadian Med. Biol. Conf.
(CMBEC30), Toronto, ON, Canada, Jun. 2007.
15) E. D. Guestrin and Eizenman M. Remote point-of-gaze estimation requiring a
single-point
calibration for applications with infants. Proceedings of ETRA 2008, pages 267-
274, Mar.
2008.
16) Hannula, D.E., Althoff, R.R, Warren, D.E., Riggs, L., Cohen N.J., and Ryan
J.D., (2010).
Worth a glance: using eye movements to investigate the cognitive neuroscience
of
memory. Frontiers in human Neuroscience, 4:Article 166, 1-13.
17) Hermans, D., Vansteenwegen, D., Eelen, P., (1999). Eye Movement
Registration as a
Continuous Index of Attention Deployment: Data from a Group of Spider Anxious
Students. Cognition & Emotion, 13(4):419-34.
18) Jansen, A., Nederkoom, C., Mulkens, S, (2005). Selective visual attention
for ugly and
beautiful body parts in eating disorders. Behavior Research and Therapy,
43(2):183-96.
19) Johansson, L., Ghaderi, A., Andersson, G., (2005). Stroop interference for
food- and body-
related words: a meta-analysis. Eating Behaviors 6(3):271-281.
20) Kellough, J.L., Beevers, C.G., Ellis, A.J., Wells, T.T., "Time course of
selective attention in
clinically depressed young adults: An eye tracking study," Behaviour Research
and
Therapy, Vol. 46, No. 11, Pp. 1238-1243, 2008.
21) Lang, P.J., Bradley, M.M., Cuthbert, B.N., 1999. International Affective
Picture System:
Instruction Manual and Affective Ratings. Technical Report A-4, University of
Florida-The
Center for Research in Psychphysiology, Gainsville, FL.
22) Long, C.O., Hinton, Cõ Gillespie, N.K., (1994). Selective processing of
food and body size
words: application of the Stroop Test with obese restrained eaters, anorexics,
and
normals. International Journal of Eating Disorders. 15(3),279-83.
23) McKhann G, Drachman D, Folstein M, et at. (1984), Clinical diagnosis of
Alzheimer's
disease: report of the NINCDS-ADRDA Work Group under the auspices of
Department of
Health and Human Services Task Force on Alzheimer's Disease. Neurology. 34
(7): 939-
44.
29
8778843.4
CA 02833398 2013-11-20
24) Mathews, A., and MacLeod, C,, (1994). Cognitive approaches to emotion and
emotional
disorders. Annual Review of Psychology, 45, 25-50.
25) Pinhas, L., & Bondy, S. J. (2010). Epidemiology of eating disorders in
Children and
Adolescents. In J. Lock (Ed.), The Oxford Handbook of Child and Adolescent
Eating
Disorders: Developmental Perspectives (Vol. 1). New York: Oxford University
Press.
26) Preti, A., Girolamo, G., Vilagut, G., Alonso, J., Graaf, R., Bruffaerts,
R., et al. (2009). The
epidemiology of eating disorders in six European countries: results of the
ESEMeD-WMH
project. Journal of Psychiatric Research, 43(14), 1125-1132.
27) Rinck, M, Becker E.S., (2006). Spider fearful individuals attend to
threat, then quickly
avoid it: evidence from eye movements. Journal of Abnormal Psychology
115(2):231-8.
28) Robert P, Onyike CU, Leentjens AF, et at, (2009), Proposed diagnostic
criteria for apathy
in Alzheimer's disease and other neuropsychiatric disorders. Eur Psychiatry.
Mar, 24 (2):
98-104.
29) Sears, C.R., Thomas C.L., Lehuquet J.M., Johnson C.S., (2010). Attentional
biases in
dysphoria: An eye-tracking study of the allocation and disengagement of
attention.
Cognition and emotion, 24(8),1349-1368.
30) Shafran R., Cooper Z., and Fairburn
(2002). Clinical perfectionism: a cognitive-
behavioural analysis. Behaviour Research and Therapy, 40, 773-791.
31) Shafran, R., Lee, M., Cooper, Z., Palmer, R.L., Fairburn, C.G., (2007).
Attentional bias in
eating disorders. International Journal of Eating Disorders, 40(4):369-80.
32) Shafran, R., Lee, M., Cooper, Z., Palmer, R.L., Fairburn, C.G., (2008).
Effect of
Psychological Treatment on Attentional Bias in Eating Disorders. International
Journal of
Eating Disorders, 41(4):348-354.
33) Smeets, E', Tiggemann, M., Kemps, E, Mills J.S., Hollitt,S., Roefs, A.,
Jansen, A., (2011).
Body Checking Induces an Attentional Bias for Body-Related Cues, International
Journal
of Eating Disorders, 44:50-57.
34) Toh, WL, Rossell SL., Castle DJ, 2011, Current visual scabpath research: a
review of
investigations into the psychotic, anxiety, and mood disorders, Comprehensive
Psychiatry,
Vol. 52 PP. 567-579.
35) VVietersheim J., Kunzi, F., Hoffmann, H., Glaub, J., Rottler E., Rraue C.,
(2012). Selective
Attention of patients with Anorexia Nervosa While looking at pictures of their
own body and
bodies of others: An Exploratory Study. Psychosomatic Medicine, 74:107-113.
8178843.4
