1
Journal of Alzheimer’s Disease 40 (2014) 1–10 DOI 10.3233/JAD-141260 IOS Press
Pr es
s
Can a Virtual Reality Cognitive Training Application Fulfill a Dual Role? Using the Virtual Supermarket Cognitive Training Application as a Screening Tool for Mild Cognitive Impairment Stelios Zygourisa,∗ , Dimitrios Giakoumisb , Konstantinos Votisb , Stefanos Doumpoulakisb , Konstantinos Ntovasa , Sofia Segkoulic , Charalampos Karagiannidisd , Dimitrios Tzovarasb and Magda Tsolakia,c a 3rd
Department of Neurology, Aristotle University of Thessaloniki, Thessaloniki, Greece Technologies Institute, Centre for Research and Technology Hellas, Thessaloniki, Greece c Greek Association of Alzheimer’s Disease and Related Disorders, Thessaloniki, Greece d Department of Special Education, University of Thessaly, Volos, Greece b Information
In
Accepted 30 October 2014
Abstract. Background: Recent research advocates the potential of virtual reality (VR) applications in assessing cognitive functions highlighting the possibility of using a VR application for mild cognitive impairment (MCI) screening. Objective: The aim of this study is to investigate whether a VR cognitive training application, the virtual supermarket (VSM), can be used as a screening tool for MCI. Methods: Two groups, one of healthy older adults (n = 21) and one of MCI patients (n = 34), were recruited from day centers for cognitive disorders and administered the VSM and a neuropsychological test battery. The performance of the two groups in the VSM was compared and correlated with performance in established neuropsychological tests. At the same time, the effectiveness of a combination of traditional neuropsychological tests and the VSM was examined. Results: VSM displayed a correct classification rate (CCR) of 87.30% when differentiating between MCI patients and healthy older adults, while it was unable to differentiate between MCI subtypes. At the same time, the VSM correlates with various established neuropsychological tests. A limited number of tests were able to improve the CCR of the VSM when combined with the VSM for screening purposes. Discussion: VSM appears to be a valid method of screening for MCI in an older adult population though it cannot be used for MCI subtype assessment. VSM’s concurrent validity is supported by the large number of correlations between the VSM and established tests. It is considered a robust test on its own as the inclusion of other tests failed to improve its CCR significantly. Keywords: Aging, Alzheimer’s disease, computers, dementia, diagnosis, memory disorders, mild cognitive impairment, usercomputer interface
∗ Correspondence to: Stelios Zygouris, MSc, 3rd Department of Neurology, Aristotle University of Thessaloniki, Sevillis 6, 54632 Thessaloniki, Greece. Tel.: +30 2310 268 471; Fax: +30 2310 351 456; E-mail: [email protected].
ISSN 1387-2877/14/$27.50 © 2014 – IOS Press and the authors. All rights reserved
S. Zygouris et al. / Can a Virtual Reality Cognitive Training Application Fulfill a Dual Role?
INTRODUCTION
In
Pr es
Recently there has been an increased interest in mild cognitive impairment (MCI) and prodromal Alzheimer’s disease (AD) [1]. MCI represents a “middle ground” between normal aging and dementia. MCI patients are able to live autonomously and perform activities of daily living (ADL), but at the same time suffer from memory impairment and impairment in complex activities such as financial planning [2, 3]. Research has shown that MCI patients present deficits in executive function [4] and navigation [5] tasks. Virtual reality (VR) applications have been used widely in the field of cognitive rehabilitation and cognitive training. VR technology enables the user to enter a computer generated environment and interact with it. It allows for a high degree of ecological validity [6] and can be tailored to suit specific clinical or research applications. The engaging nature of VR tasks makes them ideal for a wide range of applications including cognitive screening and evaluation. VR technology appears especially promising when taking into consideration that older adults often exhibit a willingness to try new technologies [7] and a positive attitude toward video games and VR environments [8]. The use of VR technology for cognitive assessment could provide an assessment method that is ecologically valid, less threatening and more enjoyable for older adults compared to standard neuropsychological tests. Research has shown that performance in a virtual reality environment can be useful in distinguishing patients with MCI from healthy older adults and that subjective memory complaints (SMCs) correlate with performance in a VR test [9]. Lately there have been some attempts to use VR applications for cognitive screening and evaluation with promising results [9–11]. Still VR technology is notably absent from commercially available computerized batteries used for cognitive screening and evaluation with only one battery (CAMCI) offering a VR subtest [12, 13]. Taking into account the latest developments in MCI research and the use of computers and VR technology in the field of cognitive aging, the Virtual Supermarket (VSM) cognitive training application was designed. This application allows older adults to train important cognitive functions such as executive function, navigation, planning, and memory. Its aim is to help MCI patients improve aspects of their cognitive functioning that are necessary for autonomous living and relate to their ability to perform activities of daily living.
Drawing upon the ever increasing literature on the use of VR shopping tasks and VR supermarket environments for the assessment of executive function [14–17] and a recent study on the use of a VR supermarket in combination with the Mini-Mental State Examination (MMSE) for MCI diagnosis [11], we decided to study whether the VSM could be used on its own as a screening tool for MCI. The aim of this study is to assess the differences in performance between healthy older adults and MCI patients and detect whether there are significant group differences in performance at the first administration of the VSM cognitive training exercise. A secondary aim is to explore to what extent the performance in the VSM correlates with established neuropsychological tests used in MCI and how it compares to them in terms of diagnostic value.
s
2
MATERIALS AND METHODS Participants
Participants were recruited from a cohort of older adults from various socioeconomic backgrounds with SMCs visiting the day centers of the Greek Association of Alzheimer’s Disease and Related Disorders (GAADRD) between June and December 2013. All participants were informed about the purpose of the study and provided their consent. Diagnosis was confirmed by a neurologist after a full neurological, neuropsychological, and laboratory assessment. MCI subtypes were determined according to Petersen criteria [2, 18]. Demographic characteristics of participants are shown in Table 1. Mean age was 68.89 years ranging from 57 to 84 years. Subjects had a mean of 11.40 years of formal education ranging from 0 to 22 years of formal education. The majority of the participants were female (45 persons). The sample included 21 healthy older adults and 34 MCI patients out of which 11 had a diagnosis of single domain amnestic MCI (aMCISD) and 23 had a diagnosis of multi domain amnestic MCI (aMCI-MD). aMCI-SD patients had a mean age of 66.55 years and a mean of 10.82 years of formal education, while aMCI-MD patients had a mean age of 72.13 years and a mean of 10.04 years of formal education. No statistically significant differences were observed between the healthy and MCI groups (as well as between the healthy group and the MCI subtype groups) in age and education while a statistically significant difference in MMSE scores was observed as
S. Zygouris et al. / Can a Virtual Reality Cognitive Training Application Fulfill a Dual Role?
3
Table 1 Characteristics of the participants Male/Female Mean Age (SE) Mean Education in years (SE) Mean MMSE (SE)
Healthy
MCI
aMCI-SD
aMCI-MD
4/17 66.57 (1.2) 13.19 (0.891) 28.86 (0.242)
6/28 70.32 (1.229) 10.29 (0.731) 27.42 (0.323)
0/11 66.55 (1.713) 10.82 (1.069) 27.27 (0.675)
6/17 72.13 (1.501) 10.04 (0.963) 27.50 (0.359)
MCI, mild cognitive impairment; aMCI-SD, single domain amnestic mild cognitive impairment; aMCI-MD, multiple domain amnestic mild cognitive impairment; MMSE, Mini Mental State Examination; SE, standard error.
s
The VSM is designed to mimic daily shopping in a supermarket, one of the most common activities of daily living. Prior to engaging in the virtual experience, the patient is asked to enter her/his age, gender, occupation, and years of formal education. A further question that has to be answered concerns the patient’s selfassessment over whether s/he is experiencing memory problems or not. The patient is then asked to navigate inside the virtual supermarket and buy the items displayed on a shopping list, appearing on the upper right corner of the screen. The patient can move to various locations on the supermarket by touching the green footprints displayed in the screen and can look around in the virtual environment by scrolling left and right on the touch screen. S/he can select and purchase products by touching them. Upon purchasing all items of the list, s/he is required to locate the cash desk and proceed to pay. The payment screen is then displayed. The user must select the precise amount needed to pay for the items bought, using a selection of Euro bills and coins. There are buttons enabling the user to correct the chosen amount or return to shopping. After paying, a statistics screen follows. The displayed information includes number and quantities of correct and incorrect items bought and total completion time. The program features four levels of difficulty depending on the number and quantities of different items on the list. The VSM is aimed at training a multitude of cognitive processes namely visual and verbal memory, executive function, attention, and spatial navigation with the emphasis placed on executive function. The need of simultaneous activation of different cognitive processes makes the program challenging enough for healthy older adults and MCI patients but precludes its use with patients suffering from dementia. The shopping list is randomized on each trial to maintain the difficulty level and limit learning effects. The software allows for active exploration of the virtual environment in order to be ecologically valid and also because active exploration has been shown to enhance learning and memory [9].
Pr es
expected. The aMCI-SD and aMCI-MD groups displayed no statistically significant differences in age, education, and MMSE scores. Neuropsychological assessment
In
Participants were administered a neuropsychological test battery including the following cognitive tests: MMSE, Ray Auditory Verbal Learning Test (RAVLT), a Greek version of the “FAS” verbal fluency test, ReyOsterrieth Complex Figure Test (ROCFT), Rivermead Behavioural Memory Test (RBMT), Test of Everyday Attention (TEA) items 1, 4, and 6, and Trail Making Test part B. It also included the following functional scales: Functional Rating Scale for Symptoms of Dementia (FRSSD), Functional Cognitive Assessment Scale (FUCAS), and the Clinical Dementia Rating (CDR). Furthermore the battery included the following measures of depression, anxiety, and neuropsychiatric symptoms: Beck Anxiety Inventory (BAI), Beck Depression Inventory (BDI), Geriatric Depression Scale (BDS), and the Perceived Stress Scale (PSS). In addition to the neuropsychological battery, participants were also administered a Greek dementia screening scale and a Greek self-report questionnaire assessing early signs of cognitive decline. Virtual reality cognitive training application
The VSM application used in this study (Supplementary Figs. 1 and 2) is a simple virtual reality cognitive training program with a low degree of immersion, based on the state of the art in the field of using VR applications for cognitive assessment [10, 14, 15]. It has been developed by the Information Technologies Institute in association with the Greek Association of Alzheimer’s Disease and Related Disorders. The application runs on any tablet device with Android® operating system, whereas PC and web-based versions also exist.
S. Zygouris et al. / Can a Virtual Reality Cognitive Training Application Fulfill a Dual Role?
Administration of the cognitive training application
Correlates between the VSM variables and established MCI measures The correlation between the duration of the test and the established variables was examined through the Kendall’s tau correlation coefficient. Table 2 summarizes the significant (p < 0.05) correlates found. Since the “Correct Types”, “Correct Quantities”, “Bought Unlisted”, and “Correct Money” variables of the VSM are in essence categorical variables with values 0 or 1, we examined their relation to the established
In
Pr es
All participants underwent a training session before the administration of the program. A medium difficulty version of the program (4 items in quantities of up to 4 for each item for a total of 12 items) was administered on a tablet device with 10-inch touch screen. During the training session, participants were asked to read the instructions as presented by the application. Afterwards a psychologist demonstrated the use of the application. Participants were shown how to navigate the visual supermarket, how to buy items, how to return unwanted items, and how to pay. They were reminded to use both the green footprints and scrolling in order to navigate the environment and instructed to pay attention to the labels above the aisles. A navigation strategy was suggested namely that they should move to a central point of the supermarket and scroll around if they have difficulty locating the aisle they were looking for. During the practice session they were allowed to navigate freely and buy any items they wanted in order to train themselves in using the VR interface. During the practice session, they could ask and receive help from the psychologist administering the application. During the test session they were allowed to ask and receive information concerning the use of the interface but no further help was given.
by examining as above, the mean education level values among the user groups formed by the outcomes of test variables (t-test on education level versus each variable: p > 0.05 for all variables).
s
4
RESULTS
Effects of gender, age, and education level on the VSM variables Gender was found to have no effect on the “Correct Types”, “Correct Quantities”, “Bought Unlisted”, and “Correct Money” variables (chi-square, p > 0.05), as well as on the duration of the test (independent samples t-test on duration versus gender: t(53) = −0.091, p = 0.817). Using independent samples T-tests no effect was found for age and education in the “Correct Types”, “Correct Quantities”, “Bought Unlisted”, and “Correct Money” variables. Through a Kendall’s tau-based correlation analysis a positive relation was found (τ = 0.296, p = 0.002) between age and duration. Through a similar procedure, negative correlation was found between duration and the education level (in years) of subjects (τ = −0.306, p = 0.002). The education level of subjects was found to have no effect on the “Correct Types”, “Correct Quantities”, “Bought Unlisted”, and “Correct Money” variables of the VSM,
Table 2 Kendall’s tau-based correlations between supermarket test duration and established variables Supermarket test variable Duration
Established variable
Correlation coefficient (τ-value)
p
MMSE FUCAS CDR TRAILB TEA1a TEA1b TEA4r TEA4t RBMT1 RBMT2 ROCFT1 ROCFT3 RAVLT1 RAVLT2 RAVLT3 RAVLT4 RAVLT5 RAVLT6 BAI
−0.209∗ 0.242∗ 0.313∗∗ 0.414∗∗ −0.254∗ −0.275∗ −0.496∗∗ 0.387∗∗ −0.328∗∗ −0.329∗∗ −0.205∗ −0.296∗∗ −0.205∗ −0.257∗∗ −0.356∗∗ −0.352∗∗ −0.243∗ −0.370∗∗ −0.295∗
0.04 0.026 0.002 0 0.02 0.011 0 0.001 0.001 0.001 0.042 0.002 0.042 0.01 0 0 0.015 0 0.018
∗ p < 0.05; ∗∗ p < 0.001.
MMSE, Mini Mental State Examination; FUCAS, Functional Cognitive Assessment Scale; CDR, Clinical Dementia Rating, TRAILB, Trail Making Test part b; TEA1a, Test of Everyday Attention item 1 – one minute score; TEA1b, Test of Everyday Attention item 1 – two minutes score; TEA4r, Test of Everyday Attention item 4 – raw score; TEA4t, Test of Everyday Attention item 4 – time; RBMT1, Rivermead Behavioural Memory Test – first recall; RBMT2, Rivermead Behavioural Memory Test – second recall; ROCFT1, Rey-Osterrieth Complex Figure Test – pattern copying; ROCF3, Rey-Osterrieth Complex Figure Test – second recall; RAVLT1, Ray Auditory Verbal Learning Test – first recall; RAVLT2, Ray Auditory Verbal Learning Test – learning ability; RAVLT3, Ray Auditory Verbal Learning Test – fifth recall; RAVLT4, Ray Auditory Verbal Learning Test – total recalled words; RAVLT5, Ray Auditory Verbal Learning Test – verbal memory; RAVLT6, Ray Auditory Verbal Learning Test – sixth recall; BAI, Beck Anxiety Inventory.
S. Zygouris et al. / Can a Virtual Reality Cognitive Training Application Fulfill a Dual Role?
5
Table 3 Difference of established variable values in respect to the VSM variables outcomes Correct Types
Correct Quantities
FALSE Mean (SE)
TRUE Mean (SE)
t-value
p
MMSE TRAILB RBMT1 RAVLT2 RAVLT3 RAVLT4 Screening Scale MMSE FUCAS TRAILB RAVLT2 TRAILB TEA4t RBMT1 RBMT2 RAVLT3 RAVLT4 RAVLT6
23.50 (0.50) 558.00 (156.98) 6.75 (0.75) 2.50 (0.50) 6.50 (0.50) 27.50 (2.50) 14.56 (0.29) 26.44 (0.63) 44.56 (0.65) 331.67 (57.96) 4.22 (0.66) 20.93 (13.85) 6.25 (0.57) 13.01 (0.57) 12.28 (0.54) 10.98 (0.37) 43.64 (1.42) 8.48 (0.53)
28.15 (0.21) 211.48 (91.46) 12.57 (0.53) 5.75 (0.30) 10.73 (0.34) 42.96 (1.30) 13.78 (0.16) 28.29 (0.23) 42.80 (0.22) 202.84 (12.64) 5.91 (0.33) 296.40 (53.31) 10.86 (2.92) 9.45 (1.05) 8.60 (1.22) 8.80 (0.65) 36.90 (2.87) 5.70 (0.62)
4.265 5.162 2.139 2.081 2.433 2.315 2.079 3.13 3.047 3.406 2.139 2.315 2.523 2.766 2.906 2.594 2.054 2.422
Journal of Alzheimer’s Disease 40 (2014) 1–10 DOI 10.3233/JAD-141260 IOS Press
Pr es
s
Can a Virtual Reality Cognitive Training Application Fulfill a Dual Role? Using the Virtual Supermarket Cognitive Training Application as a Screening Tool for Mild Cognitive Impairment Stelios Zygourisa,∗ , Dimitrios Giakoumisb , Konstantinos Votisb , Stefanos Doumpoulakisb , Konstantinos Ntovasa , Sofia Segkoulic , Charalampos Karagiannidisd , Dimitrios Tzovarasb and Magda Tsolakia,c a 3rd
Department of Neurology, Aristotle University of Thessaloniki, Thessaloniki, Greece Technologies Institute, Centre for Research and Technology Hellas, Thessaloniki, Greece c Greek Association of Alzheimer’s Disease and Related Disorders, Thessaloniki, Greece d Department of Special Education, University of Thessaly, Volos, Greece b Information
In
Accepted 30 October 2014
Abstract. Background: Recent research advocates the potential of virtual reality (VR) applications in assessing cognitive functions highlighting the possibility of using a VR application for mild cognitive impairment (MCI) screening. Objective: The aim of this study is to investigate whether a VR cognitive training application, the virtual supermarket (VSM), can be used as a screening tool for MCI. Methods: Two groups, one of healthy older adults (n = 21) and one of MCI patients (n = 34), were recruited from day centers for cognitive disorders and administered the VSM and a neuropsychological test battery. The performance of the two groups in the VSM was compared and correlated with performance in established neuropsychological tests. At the same time, the effectiveness of a combination of traditional neuropsychological tests and the VSM was examined. Results: VSM displayed a correct classification rate (CCR) of 87.30% when differentiating between MCI patients and healthy older adults, while it was unable to differentiate between MCI subtypes. At the same time, the VSM correlates with various established neuropsychological tests. A limited number of tests were able to improve the CCR of the VSM when combined with the VSM for screening purposes. Discussion: VSM appears to be a valid method of screening for MCI in an older adult population though it cannot be used for MCI subtype assessment. VSM’s concurrent validity is supported by the large number of correlations between the VSM and established tests. It is considered a robust test on its own as the inclusion of other tests failed to improve its CCR significantly. Keywords: Aging, Alzheimer’s disease, computers, dementia, diagnosis, memory disorders, mild cognitive impairment, usercomputer interface
∗ Correspondence to: Stelios Zygouris, MSc, 3rd Department of Neurology, Aristotle University of Thessaloniki, Sevillis 6, 54632 Thessaloniki, Greece. Tel.: +30 2310 268 471; Fax: +30 2310 351 456; E-mail: [email protected].
ISSN 1387-2877/14/$27.50 © 2014 – IOS Press and the authors. All rights reserved
S. Zygouris et al. / Can a Virtual Reality Cognitive Training Application Fulfill a Dual Role?
INTRODUCTION
In
Pr es
Recently there has been an increased interest in mild cognitive impairment (MCI) and prodromal Alzheimer’s disease (AD) [1]. MCI represents a “middle ground” between normal aging and dementia. MCI patients are able to live autonomously and perform activities of daily living (ADL), but at the same time suffer from memory impairment and impairment in complex activities such as financial planning [2, 3]. Research has shown that MCI patients present deficits in executive function [4] and navigation [5] tasks. Virtual reality (VR) applications have been used widely in the field of cognitive rehabilitation and cognitive training. VR technology enables the user to enter a computer generated environment and interact with it. It allows for a high degree of ecological validity [6] and can be tailored to suit specific clinical or research applications. The engaging nature of VR tasks makes them ideal for a wide range of applications including cognitive screening and evaluation. VR technology appears especially promising when taking into consideration that older adults often exhibit a willingness to try new technologies [7] and a positive attitude toward video games and VR environments [8]. The use of VR technology for cognitive assessment could provide an assessment method that is ecologically valid, less threatening and more enjoyable for older adults compared to standard neuropsychological tests. Research has shown that performance in a virtual reality environment can be useful in distinguishing patients with MCI from healthy older adults and that subjective memory complaints (SMCs) correlate with performance in a VR test [9]. Lately there have been some attempts to use VR applications for cognitive screening and evaluation with promising results [9–11]. Still VR technology is notably absent from commercially available computerized batteries used for cognitive screening and evaluation with only one battery (CAMCI) offering a VR subtest [12, 13]. Taking into account the latest developments in MCI research and the use of computers and VR technology in the field of cognitive aging, the Virtual Supermarket (VSM) cognitive training application was designed. This application allows older adults to train important cognitive functions such as executive function, navigation, planning, and memory. Its aim is to help MCI patients improve aspects of their cognitive functioning that are necessary for autonomous living and relate to their ability to perform activities of daily living.
Drawing upon the ever increasing literature on the use of VR shopping tasks and VR supermarket environments for the assessment of executive function [14–17] and a recent study on the use of a VR supermarket in combination with the Mini-Mental State Examination (MMSE) for MCI diagnosis [11], we decided to study whether the VSM could be used on its own as a screening tool for MCI. The aim of this study is to assess the differences in performance between healthy older adults and MCI patients and detect whether there are significant group differences in performance at the first administration of the VSM cognitive training exercise. A secondary aim is to explore to what extent the performance in the VSM correlates with established neuropsychological tests used in MCI and how it compares to them in terms of diagnostic value.
s
2
MATERIALS AND METHODS Participants
Participants were recruited from a cohort of older adults from various socioeconomic backgrounds with SMCs visiting the day centers of the Greek Association of Alzheimer’s Disease and Related Disorders (GAADRD) between June and December 2013. All participants were informed about the purpose of the study and provided their consent. Diagnosis was confirmed by a neurologist after a full neurological, neuropsychological, and laboratory assessment. MCI subtypes were determined according to Petersen criteria [2, 18]. Demographic characteristics of participants are shown in Table 1. Mean age was 68.89 years ranging from 57 to 84 years. Subjects had a mean of 11.40 years of formal education ranging from 0 to 22 years of formal education. The majority of the participants were female (45 persons). The sample included 21 healthy older adults and 34 MCI patients out of which 11 had a diagnosis of single domain amnestic MCI (aMCISD) and 23 had a diagnosis of multi domain amnestic MCI (aMCI-MD). aMCI-SD patients had a mean age of 66.55 years and a mean of 10.82 years of formal education, while aMCI-MD patients had a mean age of 72.13 years and a mean of 10.04 years of formal education. No statistically significant differences were observed between the healthy and MCI groups (as well as between the healthy group and the MCI subtype groups) in age and education while a statistically significant difference in MMSE scores was observed as
S. Zygouris et al. / Can a Virtual Reality Cognitive Training Application Fulfill a Dual Role?
3
Table 1 Characteristics of the participants Male/Female Mean Age (SE) Mean Education in years (SE) Mean MMSE (SE)
Healthy
MCI
aMCI-SD
aMCI-MD
4/17 66.57 (1.2) 13.19 (0.891) 28.86 (0.242)
6/28 70.32 (1.229) 10.29 (0.731) 27.42 (0.323)
0/11 66.55 (1.713) 10.82 (1.069) 27.27 (0.675)
6/17 72.13 (1.501) 10.04 (0.963) 27.50 (0.359)
MCI, mild cognitive impairment; aMCI-SD, single domain amnestic mild cognitive impairment; aMCI-MD, multiple domain amnestic mild cognitive impairment; MMSE, Mini Mental State Examination; SE, standard error.
s
The VSM is designed to mimic daily shopping in a supermarket, one of the most common activities of daily living. Prior to engaging in the virtual experience, the patient is asked to enter her/his age, gender, occupation, and years of formal education. A further question that has to be answered concerns the patient’s selfassessment over whether s/he is experiencing memory problems or not. The patient is then asked to navigate inside the virtual supermarket and buy the items displayed on a shopping list, appearing on the upper right corner of the screen. The patient can move to various locations on the supermarket by touching the green footprints displayed in the screen and can look around in the virtual environment by scrolling left and right on the touch screen. S/he can select and purchase products by touching them. Upon purchasing all items of the list, s/he is required to locate the cash desk and proceed to pay. The payment screen is then displayed. The user must select the precise amount needed to pay for the items bought, using a selection of Euro bills and coins. There are buttons enabling the user to correct the chosen amount or return to shopping. After paying, a statistics screen follows. The displayed information includes number and quantities of correct and incorrect items bought and total completion time. The program features four levels of difficulty depending on the number and quantities of different items on the list. The VSM is aimed at training a multitude of cognitive processes namely visual and verbal memory, executive function, attention, and spatial navigation with the emphasis placed on executive function. The need of simultaneous activation of different cognitive processes makes the program challenging enough for healthy older adults and MCI patients but precludes its use with patients suffering from dementia. The shopping list is randomized on each trial to maintain the difficulty level and limit learning effects. The software allows for active exploration of the virtual environment in order to be ecologically valid and also because active exploration has been shown to enhance learning and memory [9].
Pr es
expected. The aMCI-SD and aMCI-MD groups displayed no statistically significant differences in age, education, and MMSE scores. Neuropsychological assessment
In
Participants were administered a neuropsychological test battery including the following cognitive tests: MMSE, Ray Auditory Verbal Learning Test (RAVLT), a Greek version of the “FAS” verbal fluency test, ReyOsterrieth Complex Figure Test (ROCFT), Rivermead Behavioural Memory Test (RBMT), Test of Everyday Attention (TEA) items 1, 4, and 6, and Trail Making Test part B. It also included the following functional scales: Functional Rating Scale for Symptoms of Dementia (FRSSD), Functional Cognitive Assessment Scale (FUCAS), and the Clinical Dementia Rating (CDR). Furthermore the battery included the following measures of depression, anxiety, and neuropsychiatric symptoms: Beck Anxiety Inventory (BAI), Beck Depression Inventory (BDI), Geriatric Depression Scale (BDS), and the Perceived Stress Scale (PSS). In addition to the neuropsychological battery, participants were also administered a Greek dementia screening scale and a Greek self-report questionnaire assessing early signs of cognitive decline. Virtual reality cognitive training application
The VSM application used in this study (Supplementary Figs. 1 and 2) is a simple virtual reality cognitive training program with a low degree of immersion, based on the state of the art in the field of using VR applications for cognitive assessment [10, 14, 15]. It has been developed by the Information Technologies Institute in association with the Greek Association of Alzheimer’s Disease and Related Disorders. The application runs on any tablet device with Android® operating system, whereas PC and web-based versions also exist.
S. Zygouris et al. / Can a Virtual Reality Cognitive Training Application Fulfill a Dual Role?
Administration of the cognitive training application
Correlates between the VSM variables and established MCI measures The correlation between the duration of the test and the established variables was examined through the Kendall’s tau correlation coefficient. Table 2 summarizes the significant (p < 0.05) correlates found. Since the “Correct Types”, “Correct Quantities”, “Bought Unlisted”, and “Correct Money” variables of the VSM are in essence categorical variables with values 0 or 1, we examined their relation to the established
In
Pr es
All participants underwent a training session before the administration of the program. A medium difficulty version of the program (4 items in quantities of up to 4 for each item for a total of 12 items) was administered on a tablet device with 10-inch touch screen. During the training session, participants were asked to read the instructions as presented by the application. Afterwards a psychologist demonstrated the use of the application. Participants were shown how to navigate the visual supermarket, how to buy items, how to return unwanted items, and how to pay. They were reminded to use both the green footprints and scrolling in order to navigate the environment and instructed to pay attention to the labels above the aisles. A navigation strategy was suggested namely that they should move to a central point of the supermarket and scroll around if they have difficulty locating the aisle they were looking for. During the practice session they were allowed to navigate freely and buy any items they wanted in order to train themselves in using the VR interface. During the practice session, they could ask and receive help from the psychologist administering the application. During the test session they were allowed to ask and receive information concerning the use of the interface but no further help was given.
by examining as above, the mean education level values among the user groups formed by the outcomes of test variables (t-test on education level versus each variable: p > 0.05 for all variables).
s
4
RESULTS
Effects of gender, age, and education level on the VSM variables Gender was found to have no effect on the “Correct Types”, “Correct Quantities”, “Bought Unlisted”, and “Correct Money” variables (chi-square, p > 0.05), as well as on the duration of the test (independent samples t-test on duration versus gender: t(53) = −0.091, p = 0.817). Using independent samples T-tests no effect was found for age and education in the “Correct Types”, “Correct Quantities”, “Bought Unlisted”, and “Correct Money” variables. Through a Kendall’s tau-based correlation analysis a positive relation was found (τ = 0.296, p = 0.002) between age and duration. Through a similar procedure, negative correlation was found between duration and the education level (in years) of subjects (τ = −0.306, p = 0.002). The education level of subjects was found to have no effect on the “Correct Types”, “Correct Quantities”, “Bought Unlisted”, and “Correct Money” variables of the VSM,
Table 2 Kendall’s tau-based correlations between supermarket test duration and established variables Supermarket test variable Duration
Established variable
Correlation coefficient (τ-value)
p
MMSE FUCAS CDR TRAILB TEA1a TEA1b TEA4r TEA4t RBMT1 RBMT2 ROCFT1 ROCFT3 RAVLT1 RAVLT2 RAVLT3 RAVLT4 RAVLT5 RAVLT6 BAI
−0.209∗ 0.242∗ 0.313∗∗ 0.414∗∗ −0.254∗ −0.275∗ −0.496∗∗ 0.387∗∗ −0.328∗∗ −0.329∗∗ −0.205∗ −0.296∗∗ −0.205∗ −0.257∗∗ −0.356∗∗ −0.352∗∗ −0.243∗ −0.370∗∗ −0.295∗
0.04 0.026 0.002 0 0.02 0.011 0 0.001 0.001 0.001 0.042 0.002 0.042 0.01 0 0 0.015 0 0.018
∗ p < 0.05; ∗∗ p < 0.001.
MMSE, Mini Mental State Examination; FUCAS, Functional Cognitive Assessment Scale; CDR, Clinical Dementia Rating, TRAILB, Trail Making Test part b; TEA1a, Test of Everyday Attention item 1 – one minute score; TEA1b, Test of Everyday Attention item 1 – two minutes score; TEA4r, Test of Everyday Attention item 4 – raw score; TEA4t, Test of Everyday Attention item 4 – time; RBMT1, Rivermead Behavioural Memory Test – first recall; RBMT2, Rivermead Behavioural Memory Test – second recall; ROCFT1, Rey-Osterrieth Complex Figure Test – pattern copying; ROCF3, Rey-Osterrieth Complex Figure Test – second recall; RAVLT1, Ray Auditory Verbal Learning Test – first recall; RAVLT2, Ray Auditory Verbal Learning Test – learning ability; RAVLT3, Ray Auditory Verbal Learning Test – fifth recall; RAVLT4, Ray Auditory Verbal Learning Test – total recalled words; RAVLT5, Ray Auditory Verbal Learning Test – verbal memory; RAVLT6, Ray Auditory Verbal Learning Test – sixth recall; BAI, Beck Anxiety Inventory.
S. Zygouris et al. / Can a Virtual Reality Cognitive Training Application Fulfill a Dual Role?
5
Table 3 Difference of established variable values in respect to the VSM variables outcomes Correct Types
Correct Quantities
FALSE Mean (SE)
TRUE Mean (SE)
t-value
p
MMSE TRAILB RBMT1 RAVLT2 RAVLT3 RAVLT4 Screening Scale MMSE FUCAS TRAILB RAVLT2 TRAILB TEA4t RBMT1 RBMT2 RAVLT3 RAVLT4 RAVLT6
23.50 (0.50) 558.00 (156.98) 6.75 (0.75) 2.50 (0.50) 6.50 (0.50) 27.50 (2.50) 14.56 (0.29) 26.44 (0.63) 44.56 (0.65) 331.67 (57.96) 4.22 (0.66) 20.93 (13.85) 6.25 (0.57) 13.01 (0.57) 12.28 (0.54) 10.98 (0.37) 43.64 (1.42) 8.48 (0.53)
28.15 (0.21) 211.48 (91.46) 12.57 (0.53) 5.75 (0.30) 10.73 (0.34) 42.96 (1.30) 13.78 (0.16) 28.29 (0.23) 42.80 (0.22) 202.84 (12.64) 5.91 (0.33) 296.40 (53.31) 10.86 (2.92) 9.45 (1.05) 8.60 (1.22) 8.80 (0.65) 36.90 (2.87) 5.70 (0.62)
4.265 5.162 2.139 2.081 2.433 2.315 2.079 3.13 3.047 3.406 2.139 2.315 2.523 2.766 2.906 2.594 2.054 2.422
