Object recognition testing tools and techniques for measuring cognitive ability and cognitive impairment   

The present application claims the benefit of U.S. Provisional Patent Application No. 60/928,577, entitled “Computer Software-Oriented Tools and Techniques for Measuring Cognitive Ability and Cognitive Impairment,” filed May 9, 2007, the disclosure of which is incorporated by reference.

The inventions described in this patent application were made in part by government support under NIH Grant # P30 AG08017. The United States Government may have rights in these inventions.

The present disclosure relates to tools and techniques for measuring cognitive ability and/or impairment.

Many environmental and intrinsic factors influence cognitive function. Intrinsic factors that can influence cognitive function include sex, age and genetic makeup.

Effects of sex on cognitive function have been shown in humans and animal models using established tests. Sex differences have been demonstrated in both episodic memory tasks (favoring women) and spatial visualization tasks (favoring men). Interestingly, in some studies alcohol consumption abolished sex differences in spatial visualization, but not episodic memory performance. In addition, stress has been shown to differentially affect fear conditioning in men and women.

Consistent with the human studies, effects of sex on cognitive function have also been reported in animal models using established tests. In general, studies of spatial learning and memory in rodents have shown that males learn more quickly than females and exhibit superior performance in a variety of mazes. Some studies, however, have not shown such differences between the sexes. Sex differences in classical fear conditioning and shuttlebox avoidance conditioning in rats have also been reported. In addition, in some studies neonatal isolation facilitated appetitive response learning in adult female, but not male, rats.

Cognitive tests administered to humans and animals frequently involve large differences. Therefore, it often remains difficult to directly compare results on these tests across species. For example, while spatial learning and memory can be easily assessed in humans and animal models, to compare assessments of spatial learning and memory in humans and mice, navigation to a target can be important. In some tests of spatial memory, when all the information is within one field of view, the participant has an aerial perspective and a body-centered (egocentric) frame of reference (e.g. table-top tests of spatial memory). Such tests are routinely used to assess visuospatial memory, but are very different from tests of spatial memory typically used for rodents. Testing visuospatial memory in rodents typically involves a viewer perspective of a world-centered (allocentric) frame of reference with information found throughout a complex environment in which the participant has to navigate. Making direct inferences about performance on navigation tests from performance on table-top tests can be problematic.

Virtual reality (“VR”), which has been used to assess, expose, and desensitize (in phobias) event and place-related memories, to assess and teach driving and flying skills, and to distract in pain management, can also be used to assess spatial learning and memory in humans using a navigational task. Navigation in a virtual environment has been shown to be sensitive to effects of sex of participants in some, but not all, studies. In one study, a virtual environment consisting of a series of interconnected hallways, some leading to dead ends and others leading to a designated goal location, was used to study age and sex differences in spatial navigation. In this study, there was no significant effect of sex on time to complete the maze or total distance traveled, but there was an effect of sex on total number of deviations from the correct route into a dead-end corridor, and there was an effect of sex on how often participants traveled on a portion of the correct route through which they had already traveled. However, as there was no cued version of this test, it is difficult to distinguish task learning performance from spatial learning and memory performance. In another study from the same authors, a virtual water maze environment was used to study the effects of age and sex on spatial learning and memory in humans. (The water maze paradigm is commonly used to assess spatial learning and memory in rodents.) An effect of age, but not of sex, was detected on performance. In this study, a trial with a visible target was given following the trials with a hidden target.

The three major human isoforms of apolipoprotein E (APOE), which are encoded by distinct APOE alleles (ε2, ε3, and ε4), are involved in the metabolism and redistribution of lipoproteins and cholesterol. Compared with ε2 and ε3, ε4 is associated with increased risk of cognitive impairments and of developing Alzheimer\'s disease (AD). Women are at higher risk to develop AD than men, particularly women carrying ε4. In contrast to the risk to develop AD, the effects of ε4 on cognitive function in the non-demented elderly old-old (>75 years of age) are less clear. While some studies have shown poor cognitive performance in non-demented elderly ε4 carriers compared with non-demented elderly non-ε4 carriers and a small effect was observed in a meta-analysis, other studies did not.

In the elderly, high cortisol and low testosterone levels might contribute to reduced cognitive function. In older men and women, higher cortisol levels have been associated with poorer cognitive performance in some studies. However, in another study cortisol levels only inversely correlated with paragraph recall in older participants with mild cognitive impairment (MCI) but not in elderly control participants. APOE genotype might also influence cortisol levels. In AD patients, higher cerebrospinal cortisol levels in ε4 than non-ε4 carriers have been reported, although comparable cerebrospinal cortisol in non-ε4 and ε4 carriers have also been reported. In elderly men, low testosterone levels might also contribute to reduced cognitive function. In older men, testosterone levels have been positively correlated with cognitive function, and cognitive function could be improved by testosterone treatments. Similarly, testosterone, but not estrogen, levels in serum have correlated positively with cognitive performance in older women, and androgen therapy has been shown to improve cognition in surgically menopausal women. The relationship between testosterone levels and cognitive function might be ε4-dependent. In men, low testosterone levels and ε4 have been shown to interact in increasing the risk of developing AD. In addition, an interaction between ε4 and cognitive performance in healthy older men has been reported; while in non-ε4 carriers higher testosterone levels were associated with better general cognition, in ε4 carriers higher testosterone levels were associated with lower cognitive performance.

SUMMARY

This Summary is provided to introduce a selection of concepts in a simplified form that are further described below in the Detailed Description. This Summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used to limit the scope of the claimed subject matter.

In summary, the Detailed Description is directed to various techniques and tools for measuring cognitive ability and/or detecting cognitive impairment or decline. For example, techniques and tools are described that can be used to diagnose or test susceptibility to cognitive impairments in children or in elderly people (such as cognitive impairments associated with Alzheimer\'s Disease). Techniques and tools are described that can be used to evaluate treatment effects and/or measure cognitive decline over time.

The foregoing and other objects, features, and advantages of the invention will become more apparent from the following detailed description, which proceeds with reference to the accompanying figures.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a block diagram of a suitable computing environment in conjunction with which several described embodiments may be implemented.

FIG. 2 is a flowchart of a generalized technique for analysis of cognitive status using VR testing.

FIG. 3 is a flowchart of a generalized technique for analysis of cognitive status using Novel Image Novel Location (“NINL”) testing.

FIG. 4 is a diagram showing example panels of a NINL software tool according to one or more described embodiments.

FIG. 5 is a diagram showing screen shots of a virtual reality, spatial navigation software tool according to one or more described embodiments.

FIGS. 6A and 6B are charts showing comparable facial recognition scores in male and female participants, and correlation of the Faces I and Faces II scores, respectively.

FIGS. 7A-7F are charts showing NINL total scores of male and female participants, correlation of NINL I and NINL II, NINL scores for ability to detect a change, NINL scores for ability to detect a novel image, NINL scores for ability to detect a novel location of a familiar image, and correlation of combined NINL total scores and combined facial recognition total scores, respectively, according to one or more described embodiments.

FIGS. 8A-8E are charts showing, for males and females tested with a virtual reality, spatial navigation software tool in hidden target trials and visible target trials, results for latency to reach the target with (+) or without (−) wearing a head-mounted display (“HMD”), velocity, latency to reach the target, percentage time in the target quadrant, percentage of successful trials, respectively, according to one or more described embodiments. FIG. 8F is a chart showing, for males and females in a probe trial, percentage time in four quadrants, according to one or more described embodiments.

FIGS. 9A-9C are charts showing correlation between NINL total scores and latency to reach the target during a visible target session with a virtual reality, spatial navigation software tool, correlation between NINL total scores and latency to reach the target during a hidden target session with a virtual reality, spatial navigation software tool, and correlation between NINL total scores and percentage of time spent in the target quadrant during a probe trial, respectively, according to one or more described embodiments.

FIGS. 10A-10D are charts showing, for elderly women and men, an effect of sex on “Family Pictures” test scores, effect of APOE ε4 on NINL total scores (combined immediate and delayed scores), effect of APOE ε4 on Novel Location sub-scores (combined immediate and delayed scores), and effect of sex on Novel Location sub-scores, respectively, according to one or more described embodiments.

FIGS. 11A and 11B are charts showing, for elderly women and men tested with a virtual reality, spatial navigation software tool, effect of sex on velocity during visible target trials, and effect of sex on velocity during hidden target trials, respectively, according to one or more described embodiments.

FIGS. 12A-12F are charts showing, for elderly women and men tested with a virtual reality, spatial navigation software tool, effect of APOE ε4 on velocity during a visible target session, effect of ε4 on velocity during a hidden target session, effect of APOE ε4 on cumulative distance during a visible target session, effect of APOE ε4 on cumulative distance during a hidden target session, effect of APOE ε4 on latency to reach target during a visible target session, and effect of APOE ε4 on latency to reach target during a hidden target session, respectively, according to one or more described embodiments.

FIGS. 13A and 13B are charts showing, for elderly women and men tested with a virtual reality, spatial navigation software tool in a probe trial, effect of sex on percentage of time spent in four quadrants, and effect of APOE ε4 on percentage of time spent in four quadrants, respectively, according to one or more described embodiments.

FIG. 14 is a chart showing, for elderly women and men, effect of ε4 on salivary testosterone levels.

FIGS. 15A and 15B are charts showing, for elderly men, correlation of salivary cortisol levels with NINL I novel image recognition, and correlation of salivary cortisol levels with NINL II novel image recognition, respectively, according to one or more described embodiments.

FIGS. 16A, 16B and 16C are charts showing effects of APOE ε4 on performance in 7-10 year-old boys and girls tested with a virtual reality, spatial navigation software tool, according to one or more described embodiments.

DETAILED DESCRIPTION

Described embodiments are directed to techniques and tools for measuring cognitive ability and/or detecting cognitive impairment or decline. For example, techniques and tools are described that can be used to diagnose or test susceptibility to cognitive impairments in children or in elderly people (such as cognitive impairments associated with Alzheimer\'s Disease). Techniques and tools are described that can be used to evaluate treatment effects and/or measure cognitive decline over time. The various techniques and tools described herein may be used independently. Some of the described techniques and tools can be used in combination.

The following paragraphs include a discussion of terms used herein.

“Adjusting for effects of X” refers to adjusting an interpreted result such that the condition X does not skew the interpreted result.

“Age-related cognitive decline” refers to a reduction in cognition associated with advancing age, e.g., an age-related dementia.

“Artificial intelligence” refers to information processing performed by one or more computers that mimics human reasoning.

“Based at least in part on X” means based on X and zero or more other acts, results, or conditions.

“Cognitive status” refers to status of cognition—mental processes related to knowing, thinking, learning and/or judging.

“First-person” refers to a simulation of a perspective a user would have if the user were physically present in a virtual environment.

“Learning of navigation skills” refers to gradual improvement of navigation skills through repetition, such as navigation skills used in a virtual reality environment.

“Learning of landscape” refers to gradual improvement of knowledge of a landscape, such as a landscape in a virtual reality environment.

“Measuring neural activity” refers to detecting activated brain regions. For example, neural activity can be measured using functional magnetic resonance imaging (“fMRI”) techniques that measure changes in neuroanatomical activity, such as increased blood flow to areas of the brain having corresponding neurological functions.

“No change score” refers to a performance signifier that measures performance of a user in identifying situations with no change in image content or image location for a second set of one or more images relative to a first set of one or more images.

“Novel image” refers to a new image in a second set of one or more images relative to a first set of one or more images.

“Novel location” refers to a new location of an image in a second set of one or more images relative to a first set of one or more images.

“Pattern recognition” refers to identification of a pattern in data and association of the identified pattern with a condition or other data.

“Pediatric cognitive disability” refers to diminished cognition (as compared to unaffected normal peers) in a human child under the age of 18. Pediatric cognitive disability can be associated with, for example, a genetic abnormality or a neuropsychological disturbance.

“Pre-clinical Alzheimer\'s disease” refers to Alzheimer\'s disease in its early stages before memory disturbance significantly interferes with psychosocial function to an extent that a clinical diagnosis can be made based on the memory disturbance.

“Providing a treatment regimen” refers to setting or adjusting a treatment regimen, such as a dose of an anti-Alzheimer\'s disease medication or the degree to which an environment is structured to address the effects of dementia.

“Score” refers to a performance signifier, such as a number or percentage of successful trials.

“User” refers to a human being that uses or interacts with computer software and/or a computerized system.

“Virtual reality environment” refers to an environment that simulates a physical environment. For example, a computer can display a virtual reality environment to a user via a graphical display, and the user can interact with the virtual reality environment by transmitting input to the computer.

FIG. 1 illustrates a generalized example of a suitable computing environment (100) in which several of the described embodiments may be implemented. The computing environment (100) is not intended to suggest any limitation as to scope of use or functionality, as the techniques and tools may be implemented in diverse general-purpose or special-purpose computing environments.

With reference to FIG. 1, the computing environment (100) includes at least one processing unit (110) and memory (120). In FIG. 1, this most basic configuration (130) is included within a dashed line. The processing unit (110) executes computer-executable instructions and may be a real or a virtual processor. In a multi-processing system, multiple processing units execute computer-executable instructions to increase processing power. The memory (120) may be volatile memory (e.g., registers, cache, RAM), non-volatile memory (e.g., ROM, EEPROM, flash memory, etc.), or some combination of the two. The memory (120) stores software (180) implementing one or more of the described techniques and tools for testing cognitive ability and/or cognitive impairment.

A computing environment may have additional features. For example, the computing environment (100) includes storage (140), one or more input devices (150), one or more output devices (160), and one or more communication connections (170). An interconnection mechanism (not shown) such as a bus, controller, or network interconnects the components of the computing environment (100). Typically, operating system software (not shown) provides an operating environment for other software executing in the computing environment (100), and coordinates activities of the components of the computing environment (100).

The storage (140) may be removable or non-removable, and includes magnetic disks, magnetic tapes or cassettes, CD-ROMs, DVDs, flash memory, or any other medium which can be used to store information and which can be accessed within the computing environment (100). The storage (140) stores instructions for the software (180).

The input device(s) (150) may be a touch input device such as a keyboard, mouse, pen, touch screen, or trackball, a voice input device, a scanning device, or another device that provides input to the computing environment (100). For audio or video encoding, the input device(s) (150) may be a sound card, video card, TV tuner card, or similar device that accepts audio or video input in analog or digital form, or a CD-ROM, CD-RW or DVD that reads audio or video samples into the computing environment (100). The output device(s) (160) may be a display, printer, speaker, CD- or DVD-writer, or another device that provides output from the computing environment (100).

The communication connection(s) (170) enable communication over a communication medium to another computing entity. The communication medium conveys information such as computer-executable instructions, audio or video input or output, or other data in a modulated data signal. A modulated data signal is a signal that has one or more of its characteristics set or changed in such a manner as to encode information in the signal. By way of example, and not limitation, communication media include wired or wireless techniques implemented with an electrical, optical, RF, infrared, acoustic, or other carrier.

The techniques and tools can be described in the general context of computer-readable media. Computer-readable media are any available media that can be accessed within a computing environment. By way of example, and not limitation, with the computing environment (100), computer-readable media include memory (120), storage (140), communication media, and combinations of any of the above.

The techniques and tools can be described in the general context of computer-executable instructions, such as those included in program modules, being executed in a computing environment on one or more target real processors or virtual processors. Generally, program modules include routines, programs, libraries, objects, classes, components, data structures, etc. that perform particular tasks or implement particular abstract data types. The functionality of the program modules may be combined or split between program modules as desired in various embodiments. Computer-executable instructions for program modules may be executed within a local or distributed computing environment.

FIG. 2 shows a generalized technique (200) for analysis of cognitive status of a user using testing with a virtual reality (“VR”) environment. A software tool such as one operating in the computer system environment shown in FIG. 1 or other tool performs the technique. Example use scenarios and example clinical applications for the generalized technique (200) are described below.

The tool receives (210) input from the user as the user interacts with software presenting a VR environment. In example implementations (including those described in the following sections), the VR environment includes a first-person, three-dimensional graphical rendering of the environment as well as sound cues for the environment. The environment is graphically rendered on a computer monitor or, for a more immersive experience, presented to the user using virtual reality goggles or another head mounted display. Inputs (such as direction of movement, speed of movement) are received from the user using a force-feedback joystick, other joystick, mouse, keyboard or other input device.

Returning to FIG. 2, the tool measures (220) performance of the user in the VR environment based at least in part upon the received input. The organization of the VR environment depends on implementation but typically includes areas such as quadrants which may be organized in terms of a coordinate space. As an intermediate part of measuring performance, for example, the tool tracks position of the user over time in the coordinate space. The results of the tracking are stored in memory or a file (e.g., as timestamped coordinate locations) for later analysis of patterns of user behavior.

In terms of metrics, the tool measures one or more of the following: (1) distance (e.g., cumulative or start to target) traversed in the VR environment, (2) time elapsed before reaching a target (or targets) in the VR environment, (3) percentage of successful trials (where success is, e.g., finding a target), (4) time spent in a target area of the VR environment, (5) velocity of movement in the VR environment, (6) pattern of movement (e.g., between multiple areas or in terms of coordinates) in the VR environment, and/or (7) pattern of time spent in respective areas of the VR environment. Alternatively, the tool measures performance using other and/or additional metrics. Some metrics (such as velocity of movement and time elapsed before reaching a target) may depend on each other to some extent, while other metrics do not.

In some implementations, the tool measures performance in a series of VR tests, with some tests having one or more “visible” targets and other tests having one or more “hidden” targets. In the “visible” target trials, one or more visual or audible cues assist the user in finding a target. For example, a prominent flag or other graphical cue is placed next to the target to help the user find the target, or directional arrows guide the user to the target. In the hidden target trials, the performance of the user in finding the target(s) is measured without giving the user the cues from the visible target testing. Such tests help measure memory retention of the user in navigating the VR environment.

Returning to FIG. 2, the tool uses (230) the measured performance of the user in the VR environment in analysis of cognitive status. For example, the tool assesses: (a) the presence or extent of age-related cognitive decline (e.g., a decline in memory performance or learning performance), (b) presence or extent of pediatric cognitive disability (e.g., a memory performance problem or learning performance problem), (c) presence or extent of progression of Alzheimer\'s disease, (d) presence of a characteristic of pre-clinical Alzheimer\'s disease, and/or (e) response of the user to therapeutic intervention to treat cognitive decline. To make the assessment, the tool can use artificial intelligence mechanisms such as classifiers (e.g., neural networks) for pattern recognition, statistical analysis, etc. Example therapeutic interventions are presented below. Alternatively, the tool uses the measured performance for a different type of analysis.

The cognitive status assessment relates the measured performance to a cognitive status classification. In making the assessment, the tool can compensate for the effects of sex, age and/or learning about the VR environment (e.g., navigation skills, landscape) on the measured performance. The following sections describe observed correlations between sex, age and learning in example uses of the generalized technique (200), and such correlations can be compensated for during the assessment of cognitive status.

In some implementations, the user repeatedly takes the VR navigation test and the performance of the user over time is measured so as to assess changes in cognitive status of the user. Typically, this involves comparing cognitive status assessments from trial to trial for the user. In other implementations, the results of testing are compared for multiple users, e.g., as part of population studies for the efficacy of a therapy.

FIG. 3 shows a generalized technique (300) for using measured performance of a user on a Novel Image Novel Location (“NINL”) test in analysis of cognitive status. A software tool such as one operating in the computer system environment shown in FIG. 1 or other tool performs the technique. Alternatively, the NINL test is administered by a human supervisor using paper materials. Example use scenarios and example clinical applications for the generalized technique (300) are described below.

The tool receives (310) input from the user as the user takes the NINL test. In example implementations, a software tool graphically presents the NINL test to the user on a computer monitor as a series of images in panels or a slideshow. The software tool accepts input from the user via a keyboard, touchpad or mouse, or the software tool receives and process voice input from the user, or the software tool receives and processes another kind of input. In a typical NINL test, the input indicates whether the user perceives “no change” in an image or group of images presented to the user, a “new location” in the image(s), or a “new image” in the image(s). Alternatively, the choices presented to the user and/or selections received from the user have a different format.

Returning to FIG. 3, the tool measures (320) performance of the user in the NINL test based at least in part upon the received input. In some implementations, as described in the following sections, the tool presents a first set of images organized by location to the user. For example, the first set of images includes multiple panels with each panel including multiple images. The tool then presents a second set of images organized by location to the user. For example, the second set of images includes multiple panels with each panel including multiple images. To measure whether the user detects changes in image content and/or location, one or more images of the second set of images differs in image content and/or image location relative to the first set of images. Alternatively, the sets of images for the NINL test have a different configuration.

In terms of metrics, the tool measures one or more of the following: (1) a novel location score indicating performance of the user in identifying changes in locations of images; (2) a novel image score indicating performance of the user in identifying new images in a second set of one or more images relative to a first set of one or more images; and/or (3) a no change score indicating performance of the user in identifying situations with no change in image content or image location for a second set of one or more images relative to a first set of one or more images. Alternatively, the tool measures performance using other and/or additional metrics.

In some implementations, the tool measures performance in an “immediate” NINL test and also measures performance in a “delayed” NINL test. For example, the immediate NINL test occurs shortly after the user reviews a training set of images for the NINL test, and the delayed NINL test occurs some defined period (e.g., five minutes) after the immediate NINL test. The duration of the defined delay period depends on implementation and is set to measure memory retention performance of the user.

Returning to FIG. 3, the tool uses (330) the measured performance of the user on the test in analysis of cognitive status. For example, the tool assesses: (a) the presence or extent of age-related cognitive decline (e.g., a decline in memory performance or learning performance), (b) presence or extent of pediatric cognitive disability (e.g., a memory performance problem or learning performance problem), (c) presence or extent of progression of Alzheimer\'s disease, (d) presence of a characteristic of pre-clinical Alzheimer\'s disease, and/or (e) response of the user to therapeutic intervention to treat cognitive decline. Example therapeutic interventions are presented below. Alternatively, the tool assesses cognitive status for a different type of cognitive assessment.

The cognitive status assessment relates the measured performance to a cognitive status classification. In making the assessment, the tool can compensate for the effects of sex, age and/or learning about the testing on the measured performance. The following sections describe observed correlations between sex, age and learning in example uses of the generalized technique (300), and such correlations can be compensated for during the assessment of cognitive status.

In some implementations, the user repeatedly takes the NINL test and the performance of the user over time is measured so as to assess changes in cognitive status of the user. Typically, this involves comparing cognitive status assessments from trial to trial for the user. In other implementations, the results of testing are compared for multiple users, e.g., as part of population studies for the efficacy of a therapy.

The generalized techniques (200, 300) can be used in various scenarios, including but not limited to home use scenarios, professional use scenarios with fMRI equipment, professional use scenarios with MRI equipment, and professional use scenarios with just the testing.

For example, when used with MRI equipment (or fMRI equipment), the equipment measures neural activity of the user as the user takes the NINL test and/or VR test. A cognitive status assessment for the user can then also be based on the measured neural activity.

Or, when used in a home use scenario, the user takes the NINL test and/or the VR test on a home computer system such as a desktop or laptop computer. The test can be delivered to the user on a computer-readable medium such as a disk or delivered to the user over a network connection from a server computer system. The user inputs can be received and processed locally to measure performance and assess cognitive status, or information can be forwarded to a remote server computer site to measure performance and/or assess cognitive status.

Alternatively, the NINL testing and/or VR testing is performed in conjunction with other and/or additional batteries of cognitive tests or physical evaluations of the user.

Described implementations can be used in a variety of contexts, such as psychological testing or clinical trials involving children and/or the elderly. The NINL object recognition test and the Memory Island spatial navigation test turned out to be sensitive to detect differences in learning and memory performance in these two populations. The sensitivity of these tests is a potential benefit or advantage over other technology. For example, in contrast to established cognitive tests, these tests were shown to be sensitive to effects of APOE ε4 (a risk factor for developing Alzheimer\'s disease and cognitive impairments following various environmental challenges) in non-demented elderly and children. Other advantages and problems to be solved are presented herein.

In example implementations, cognitive status assessments are used to make decisions about therapeutic interventions for users (e.g., children or elderly). In general, in this context, a therapeutic intervention is a known or proposed treatment for cognitive decline, such as the cognitive decline caused by normal aging or a pathological process, such as Alzheimer\'s disease or another condition associated with dementia, such as a neurological disease (for example, Huntington\'s Disease, Parkinson\'s disease, Creutzfeldt-Jakob Disease or a brain tumor), a vascular disorder (such as multi-infarct dementia or stroke), an infectious etiology (such as HIV/AIDS, spongiform encephalopathy, or syphilis), a toxic exposure (for example, to lead or alcohol), or an undesired effect of a drug. When the treatment is a proposed treatment, it can be administered as part of a clinical trial, and the response of the subject to the treatment can be assessed by the performance of the subject in the VR environment and/or NINL testing.

For example, the therapeutic intervention includes a drug therapy, and the cognitive status assessment is used in determining a therapeutic dose of the drug therapy. In the context of treating cognitive decline, for example, the therapeutic intervention is an APOE ε4 inhibitor, an APOE ε3 or APOE ε2 mimetic, a cholinesterase inhibitor, an N-methyl-aspartate receptor antagonist, or a vitamin. Or, the therapeutic intervention includes hormone therapy using testosterone and/or another androgen, or using estrogen.

This section describes example implementations of NINL tests and VR spatial navigation tests, then details results of performance on the tests in a first series of trials. It includes discussion of specific problems addressed and advantages for the example test implementations in some contexts. Alternatively, implementations of the NINL and VR spatial navigation techniques and tools vary in terms of technical details, specific advantages and/or problems solved.

A computer-generated VR island environment was developed to mirror the water maze paradigm of spatial learning and memory sensitive to effects of sex on age-related cognitive decline in mouse studies. The participants were trained to navigate to a visible target and subsequently to a hidden target. A joystick was used to control direction and speed of body movement. In addition, potential effects of sex on facial recognition and object recognition were tested using faces and NINL testing, respectively.

A. Materials and Methods

1. Participants

To determine the effects of sex on cognitive test performance, 27 community college students between 20 and 44 years of age (mean age±S.E.M., 30.3±1.2 years of age; 14 males (mean age±S.E.M., 29.1±1.4 years of age) and 13 females (mean age±S.E.M., 31.5±1.9 years of age)) were tested.

To determine whether use of a head-mounted display (“HMD”) system (HMD model V8, Virtual Research System Inc., Santa Clara, Calif.) influences performance in a spatial learning and memory test requiring navigation (see below), 24 additional young participants between 17 and 40 years of age (mean age±S.E.M., 30.5±1.2 years of age; 14 males and 10 females) were recruited from the Oregon Health Science University campus community.

2. Facial Recognition

The testing began with two non-computerized memory tests. The first of these was a facial recognition test (Faces I and Faces II), a part of the Wechsler Memory Scale III developed and published by the Psychological Corporation. In this test, the participant was shown a series of 24 faces and asked to remember each one. Immediately after, the participant was shown another series of 48 faces (the 24 original faces plus 24 distracter faces) and asked to indicate whether each face was one of the faces they were directed to remember earlier or not (Faces I score). After an interval of five minutes, the participant was shown a different set of 48 faces (the same 24 original faces plus 24 new distracter faces) and again asked to indicate whether each face was one of the faces they were directed to remember earlier or not (Faces II score). For the Faces I and Faces II scores, the participant received one point for a correct response and zero points for an incorrect response with a maximal total of 48 points.

3. Object Recognition

Following the facial recognition test, an object recognition test entitled Novel Image, Novel Location (“NINL”) test was presented to the study participants. In this test, the participant was presented with a series of 12 panels, one at a time, for eight seconds each. Each panel consisted of four quadrants (A, B, C, and D), with a different image in three of the four quadrants. The images were all similar in complexity but different in content. Positioning of the images within three of the four quadrants varied between panels.

FIG. 4 is a diagram showing example panels of the NINL software tool. On the left are panels from the first set. On the right are the corresponding panels from the second set, containing a novel location (A), novel image (B), or no change (C).

For each panel, the participant was asked to remember the images and their positions. After the participant had been presented the first set of 12 panels, they were immediately presented with a second set of 12 panels. After five minutes, they were again presented the second set of 12 panels. In the second set, the panels were either identical to, or slightly different from their counterparts in the first set. The variations in the new panels were either in the positioning of one of the three images (Novel Location) or in that they contained a novel image in the location previously containing one of the three familiar images (Novel Image). Out of the 12 panels in the second set, 4 panels were identical to panels shown in the first set, four panels contained a familiar image in a novel location, and four panels contained a novel image in place of a familiar image. For each of the 12 panels in the second set, the participant was asked to identify the new panel as being either identical to the corresponding panel in the first set (“Yes” answer), or containing a novel image or a novel location of a familiar image (“No” answer), with a maximum score of 12. These answers provided the total score.

Test performance was also analyzed with four sub-scores. If a panel was identified as containing a novel image, the participant was asked which image on the panel was novel. If a panel was identified as containing a novel location of a familiar image, the participant was asked the novel location of the familiar image on the panel. The “No Change” sub-score (four points max) reflected correct identification of panels identical to those seen in the first set. No points were deducted for incorrect identification of a panel as being identical to one seen in the first set. The “Change” sub-score (eight points max) reflected the ability to identify and characterize the type of anomaly in the panel (novel location or novel image), but not whether the particular image that changed was identified. No points were given when a change was indicated but the type of anomaly (novel location or novel image) was not correctly identified. The final two sub-scores reflected the ability to correctly identify the exact Novel Location (maximal four points) or Novel Image (maximal four points). In preliminary studies involving study participants age-matched to those in the current study, the version of the object recognition test described worked well and did not lead to a ceiling effect in test performance as a result of being too easy.

4. Memory Island

Next, a computer-generated virtual reality world (Memory Island) was used to assess spatial learning and memory. The participants were immersed in a computer-generated three-dimensional environment through a HMD system (HMD model V8, Virtual Research System Inc., Santa Clara, Calif.) comprised of special LCD video goggles and Sennheiser headphones. Inside the visor of the helmet were two video screens, one for each eye, generating a three-dimensional visual experience. Two earphones presented stereo sounds that coincided with the visual images in the visor, further enhancing the immersion experience. A Microsoft Sidewinder joystick determined the direction and speed of movement in the virtual world. As mentioned earlier, to determine whether use of the HMD influences performance in a spatial learning and memory test requiring navigation, an additional cohort of participants was tested with and without the HMD in two subsequent sessions using a counterbalanced design. Each session included four visible target trials, four hidden target trials (target only visible in very close proximity to the target), and a probe trial (no target present). Movement of the participant was tracked and recorded in time-stamped coordinate files, which were used to calculate speed of movement, time to reach the target (latency), and percentage time spent in each quadrant during the visible target session and hidden target session. Percentage time spent in each quadrant is a valuable measure, as it is usually independent of velocity.

The virtual world simulated an island environment of 347 m×287 m comprised of four quadrants. FIG. 5 is a diagram showing screen shots of the virtual reality, spatial navigation Memory Island software tool. A flag marks the location of the target during the visible target session (A) while no flag is present during the hidden target session (B). Each quadrant of the island has a different target. The target in quadrant 1 is a fountain (C), in quadrant 2 a piece of moving art (D), in quadrant 3 a seal (E), and in quadrant 4 a seagull (F).

As shown in FIG. 5, each quadrant contained a different target item. The participant was first asked to navigate to a target location visibly marked with a flag adjacent to the target (visible target). Targets in all four quadrants were used for visible target training in four consecutive trials. The starting orientation of the participant was varied in each trial, and these variations were kept consistent for all participants. As the starting orientation for a particular trial influenced the difficulty level of that trial, mean performance over the four trials of the visible or hidden target session were used for data analysis rather than performance during individual trials. After training to locate the visible targets, the participant was trained to navigate to a hidden target (here, no flag adjacent to the target, so the participant had to remember where the hidden target was and how to get there). The location of the hidden target was kept constant for each participant. Participants were given four trials with the hidden target. If the participant was unable to locate the target within two minutes, an arrow appeared to guide them to it. Trials in which the target was located within two minutes were defined as successful trials. The percentage of successful trials in the visible and hidden target session was used as an additional performance measure. Following the hidden target trials, the participant received a thirty second probe trial (target removed).

5. Statistical Analysis

Statistical differences between groups were determined by ANOVA, with sex as between participant factor, followed by Tukey-Kramer post hoc tests when appropriate. For analyzing probe trial data on Memory Island, the environment was divided into four quadrants and data was analyzed for the percentage of time spent in each quadrant, with the percentage of time spent in each quadrant as a within-participant measure. To assess significance of linear correlations, Pearson correlation calculations with two-tailed p values were used. All these statistics were performed using JMP software (SAS Institute Inc., Cary, N.C.).

B. Results of the First Series of Trials

human tests designed to mirror rodent tests of object recognition and spatial navigation were administered to adult cognitively healthy humans. Facial recognition was also assessed. The trial results showed no statistically significant sex difference in facial recognition, consistent with earlier studies. In the object recognition test, the test-retest NINL total scores during the same visit were highly correlated, comparable to the test-retest correlations obtained in the established facial recognition test. No statistically significant effects were identified for sex on object recognition. However, in the spatial navigation test, effects were identified for sex on spatial learning and memory during the session with the hidden, but not visible, target. These tests are useful to compare assessments of object recognition and spatial learning and memory in humans and animal models.

1. Facial Recognition Scores

First, facial recognition was assessed. FIGS. 6A and 6B are charts showing comparable facial recognition scores in male and female participants, and correlation of the Faces I and Faces II scores, respectively (n=14 males and n=13 females). No statistically significant effect was identified for sex (F=0.5043, p=0.6810, FIG. 6A) on facial recognition scores. The scores of Faces I and Faces II were highly correlated (r=0.8182, p<0.0001, FIG. 6B).

2. Novel Image and Novel Location (NINL)

Next, participants were tested for object recognition. FIGS. 7A-7F are charts showing (A) NINL total scores of male and female participants, (B) correlation of NINL I and NINL II, (C) scores indicating ability to detect a change, (D) scores indicating ability to detect a novel image, (E) scores indicating ability to detect a novel location of a familiar image, and (F) correlation of combined NINL total scores and combined facial recognition total scores, respectively (n=14 males and n=13 females).

As with the facial recognition test, no statistically significant effect was identified for sex on NINL total scores (F=0.5805, p=0.6305, FIG. 7A). The scores of NINL trials 1 and 2 were highly correlated (r=0.8775, p<0.0001, FIG. 7B). Interestingly, the combined total scores for facial recognition and NINL total scores were also highly correlated (r=0.5228, p<0.005, FIG. 7F).

With regard to the sub-scores, male and female participants showed no difference in their ability to detect a change (F=0.3183, p=0.8121, FIG. 7C), a novel image (F=0.6360, p=0.5952, FIG. 7D) or novel location (F=0.4148, p=0.7431, FIG. 7E).

3. Spatial Learning and Memory Requiring Navigation (Memory Island)

Finally, spatial learning and memory requiring navigation were assessed on Memory Island. FIGS. 8A-8E are charts showing, for males and females tested with a virtual reality, spatial navigation software tool in hidden target and visible target trials, (A) results for latency to reach the target with (+) or without (−) wearing a HMD, (B) velocity, (C) latency to reach the target, (D) percentage time in the target quadrant, (E) percentage of successful trials, respectively. FIG. 8F is a chart showing, for males and females in a probe trial, percentage time in four quadrants. In FIG. 8A, n=14 males and n=10 females for (A). In FIGS. 8B-8F, n=14 males and n=13 females.

The participants were first trained to locate a visible target in four trials (visible target session). Subsequently, they were trained to locate a hidden target in four trials (hidden target session). No statistically significant effect was identified for the use of the HMD to perform this task on time to locate the target (latency) (F=0.92, p=0.5150, FIG. 8A), velocity (F=1.24, p=0.5300, FIG. 8B) or percentage time spent in the target quadrant (F=2.14, p=0.323, FIG. 8D) during the visible or hidden target session.

In both the visible target session and the hidden target session, the female participants moved slower (lower velocities) (F=15.59, p<0.0002, FIG. 8B) than the male participants. Analyzing the visible and hidden target sessions combined by repeated measures, the female participants showed higher latencies (F=19.22, p<0.0001, FIG. 8C) than the male participants and there was a sex×session interaction (F=7.80, p=0.008). In the hidden target session, but not the visible target session, the females showed higher latencies than the males (FIG. 8C). As the female participants moved slower than the male participants in both the visible and hidden target session and the magnitude of this sex difference was comparable in the visible and the hidden target session (FIG. 8B), the sex difference in moving speeds did not account for the sex difference in latencies in the hidden target session (FIG. 8C).

Percentage time in the target quadrant, which is typically not affected by velocity, was also measured. In the visible target session, female and male participants had no difficulty in locating the target and spent close to 100% of their time searching in the target quadrant (FIG. 8D). In contrast, in the hidden target session, female participants spent less time in the target quadrant than male participants (F=12.27, p<0.001).

Additionally, the percentage of “successful” trials for each participant was measured (FIG. 8E). A successful trial was defined as a trial in which the target was located within 120 seconds. Female, participants had fewer successful trials than male participants in the hidden target session (F=10.29, p=0.0021), but not the visible target session (F=1.94, p=0.2652).

Following the hidden target session, the participants performed a 30-second probe trial in which there was no target present. The participants were unaware of the absence of the target during the probe trial, and were asked to perform one last trial with the hidden target. Both females and males spent most of their time searching in the target quadrant (FIG. 8F). There was a trend towards a sex difference with the males spending more time in the target quadrant than the females, but that did not reach significance (F=3.49, p=0.0715).

4. Sex and Performance

Since there were sex differences in spatial navigation measures on Memory Island, these measures were examined for correlation with performance on the other behavioral tests. FIGS. 9A-9C are charts showing correlation between NINL total scores and latency to reach the target during a visible target session with a virtual reality, spatial navigation software tool, correlation between NINL total scores and latency to reach the target during a hidden target session with a virtual reality, spatial navigation software tool, and correlation between NINL total scores and percentage of time spent in the target quadrant during a probe trial, respectively.

In female, but not male, participants the combined NINL total scores correlated with average time to reach the target during the visible target session of Memory Island (r=−0.6736, p<0.01, FIG. 9A), with average time to reach the target during the hidden target session of Memory Island (r=−0.6005, p<0.03, FIG. 9B), and with percentage of time spent in the target quadrant in the probe trial (r=0.7217, p<0.01, FIG. 9C).

D. Discussion

In the object recognition test, the test-retest NINL total scores during the same visit were highly correlated, comparable to the test-retest correlations obtained in the established facial recognition test. In the spatial navigation test, effects were identified for sex on spatial learning and memory during the session with the hidden, but not visible, target. No statistically significant effects were identified for sex on object recognition.

There was a sex difference in the percentage of time in the target quadrant during the hidden target, but not visible target, session. This measure is independent of velocity and not biased by start location, as all participants started out in the center of the island. Therefore, these data tend to show a sex difference in ability to locate the hidden target per se, rather than in general ability to perform this task regardless of whether the target was visible or hidden.

The identified sex differences in spatial learning and memory on Memory Island are consistent with sex differences in visual spatial perception and in spatial learning and memory in real and other virtual environment navigation tasks. Functional magnetic resonance imaging (“fMRI”) during navigational tasks has shown that women recruit the right parietal and right prefrontal area, whereas men recruit the left hippocampal area, which may relate to the predominant use of landmark cues by women and geometric and landmark cues by men. However, it might be more complex. The Memory Island test environment contains landmarks predicting the target location and still showed sex differences in performance. These data tend to show that the sex differences in spatial memory do not require the exclusion of stable landmarks.

The Memory Island test can be distinguished from prior studies that have adapted the water maze test to study spatial learning and memory in humans using VR. Programs designed to mirror the water maze test in rodent studies might lack elements found in real world situations. Compared to the prior studies, Memory Island involves a higher degree of immersion into the virtual environment. For example, Memory Island also contains environmental sounds (e.g., birds). While not a water maze environment, the design and analysis of the water maze test was incorporated into the design of Memory Island. Importantly, none of the participants experienced nausea or dizziness on Memory Island, while 10% of the participants experienced these symptoms after exposure to a virtual environment of interconnected hallways and other virtual environments in some prior studies.

In contrast to Memory Island, in NINL testing, no statistically significant effects were identified for sex on object recognition.

These tests are useful in comparing assessments of object recognition and spatial learning and memory in humans and animal models.

After describing example implementations of NINL tests and VR spatial navigation tests, this section details results of performance on the tests in a second series of trials. It includes discussion of specific problems addressed and advantages for the example test implementations in some contexts. Alternatively, implementations of the NINL and VR spatial navigation techniques and tools vary in terms of technical details, specific advantages and/or problems solved.

In the second series of trials using example implementations of NINL tests and VR tests, to determine effects of APOE ε4 (ε4) on cognitive performance of healthy elderly, 115 non-demented elders (mean age 81 years) were cognitive tested. The established tests Faces, Family Pictures, Spatial Span Forward and Backward, as well as the object recognition and spatial navigation tests described herein, were used as cognitive tests. Salivary samples were collected to determine APOE genotype and salivary testosterone and cortisol levels.

Non-ε4- and ε4-carrying men and women did not differ in age, or Mini-Mental State Examination, Wide Range Achievement Test-Reading, Beck Anxiety Inventory, or reaction time scores. In the second series of trials, an effect was identified for ε4 on the object recognition and spatial navigation tests, however, with non-ε4 carriers outperforming ε4 carriers, but not in the other cognitive tests. No relationship was found for sex and ε4 status or sex and performance during the hidden target session of Memory Island. In men, salivary cortisol levels correlated with object recognition. These results show that object recognition and spatial navigation tests are useful to assess cognitive function in the elderly.

A. Procedures

1. Study Participants

To determine the effects of sex and ε4 on cognitive performance in the non-demented elderly, people ranging in age from 62 to 92 (mean age±S.E.M., 81.60±0.57 years) were tested. The inclusion criteria were: 1) age 55 and over; and 2) stable medical conditions. Exclusion criteria were vision or hearing deficits severe enough to interfere with cognitive testing. Participants were given a Mini-Mental State Examination (“MMSE”), a short questionnaire that tests different areas of cognitive function, with a maximum score of 30. (See Kurlowicz et al., “The Mini Mental State Examination (MMSE),” Try This: Best Practices in Nursing Care to Older Adults, Hartford Institute for Geriatric Nursing, no. 3 (January 1999).)

All participants had MMSE scores equal or greater than 22 (see below).

The final sample was composed of 115 participants, all whites. The sample was divided into two APOE genotype groups, ε4 carriers and non-ε4 carriers. Those in the non-ε4 carriers group represented ε3/ε3 homozygotes and ε2/ε3 heterozygotes. Those in the ε4 carriers group represented ε4/ε4 homozygotes, ε2/ε4, and ε3/ε4 heterozygotes (Table 1).

TABLE 1 APOE genotype distribution of study participants. Genotype Women Men ε2/ε3 13 (15.1%) 2 (6.9%) ε2/ε4 1 (1.2%) 0 (0.0%) ε3/ε3 52 (59.3%) 22 (75.9%) ε3/ε4 18 (20.9)   5 (17.2%) ε4/ε4 2 (2.3%) 0 (0.0%) Values are presented as N (%) of women and men for each genotype.

The group of women consisted of ε6 individuals (mean age±S.E.M., 81.2±0.7 years of age), among them 65 non-ε4 carriers and 21 ε4 carriers. The group of men consisted of 29 individuals (mean age±S.E.M., 82.9±0.9 years of age), among them 24 non-ε4 carriers and five ε4 carriers. There was no significant sex difference in the proportion of ε4 carriers among men and women. When cognitive status of the participants was assessed using the MMSE, 111 participants had a MMSE score greater than 23 which corresponds to a cutoff score for cognitively healthy people. The four participants who obtained a MMSE score below 24 (three scored 23, one scored 22) performed well on the other cognitive tests. As MMSE scores can be affected by other conditions such as hearing impairment, the data were analyzed with and without these four individuals included. Both analyses revealed a similar pattern of results. Therefore, these four participants were not excluded from the study.

Premorbid intellectual functioning general intelligence levels were evaluated using the Wide Range Achievement Test-Reading (“WRAT-R”) instead of years of formal education. As anxiety levels and reaction times can influence performance on cognitive tests, they were analyzed as well. Levels of anxiety were assessed using the Beck Anxiety Inventory (“BAI”). Reaction times were measured by presenting (on a computer screen) a series of colored ellipses at varying time intervals and asking the participants to press a button as soon as the ellipse appeared (Gary Darby, “Reaction Times,” http://www.delphiforfun.org/Programs/Reaction_times.htm (©2000-2007)). The amount of time between the appearance of the stimulus and the time the button was pressed was recorded. No statistically significant differences were identified for age; cognitive status, pre-morbid intellectual functioning, anxiety levels or reaction times between men and women or non-ε4 and ε4-carrying study participants, respectively. (See Table 2.) The person testing the study participants was blinded to APOE genotype.

TABLE 2 Demography of study participants.

Download full PDF for full patent description/claims.




You can also Monitor Keywords and Search for tracking patents relating to this Object recognition testing tools and techniques for measuring cognitive ability and cognitive impairment patent application.

Patent Applications in related categories:

20130149683 - System and method for symptom based reporting - The present invention provides a computer-assisted process for receiving cognitive behavior data for use in a therapy, the computer-assisted process comprises the steps of: (a) providing at least one mobile device to a patient, wherein the at least one mobile device is functionally connected to a central repository; (b) capturing ...


###


Other recent patent applications listed under the agent :