An Investigation of Computer-based Brain Training on the Cognitive and EEG Performance of Employees

Conference Paper (PDF Available)inConference proceedings: ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Conference 2019 · July 2019with 567 Reads 
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
DOI: 10.1109/EMBC.2019.8856758 ·
Conference: 41st IEEE Engineering in Medicine and Biology Conference 2019, At Berlin, Germany
Cite this publication
Abstract
Neurocognitive skills (e.g., processing speed, attention and memory) were hypothesized to be critical for workplace performance and by extension for the work-life balance of employees. Twenty-one employee volunteers underwent a neurocognitive training program-which consisted of an initial pre-test assessment, a six week "boost" or intervention period, and then a re-assessment to track the progress of each participant. A median split of the group created two training groups: a long-training group that averaged 30 hours of total training during the boost period; and a short-training group that averaged 7 hours of training. On pre-training measures of neurocognitive performance, group differences in performance did not reach statistical significance. Following training participants experienced a positive impact from the program as measured in three ways: standardized higher behavioral metrics, improved cognitive state metrics using EEG and positive self-reported data. From a quantitative perspective, participants' cognitive efficiency increased by 12% for the high-training group and 5% for the low-training group (cognitive efficiency refers to a behavioral measure which combines accuracy and speed). Qualitatively, study participants reported improvements in their productivity and mental performance post-study.
Advertisement
AbstractNeurocognitive skills (e.g., processing speed,
attention and memory) were hypothesized to be critical for
workplace performance and by extension for the work-life
balance of employees. Twenty-one employee volunteers
underwent a neurocognitive training program which consisted
of an initial pre-test assessment, a six week “boost” or
intervention period, and then a re-assessment to track the
progress of each participant. A median split of the group created
two training groups: a long-training group that averaged 30
hours of total training during the boost period; and a short-
training group that averaged 7 hours of training. On pre-
training measures of neurocognitive performance, group
differences in performance did not reach statistical significance.
Following training participants experienced a positive impact
from the program as measured in three ways: standardized
higher behavioral metrics, improved cognitive state metrics
using EEG and positive self-reported data. From a quantitative
perspective, participantscognitive efficiency increased by 12%
for the high-training group and 5% for the low-training group
(cognitive efficiency refers to a behavioral measure which
combines accuracy and speed). Qualitatively, study participants
reported improvements in their productivity and mental
performance post-study.
I. INTRODUCTION
Measures of cognitive ability and overall brain health are
significant predictors of employment status, future cognitive
decline and brain disease [1]. In older adult populations brain
training programs have demonstrated positive benefits
immediately following brain training as well as longitudinally
including at 10-year follow-ups [2]. While results for slowing
cognitive decline in the aged are gaining acceptance among
many in the scientific community, few studies have extended
brain training approaches to improving workplace
performance or employee cognition [3].
The present study was implemented to establish the utility
of a neuroscience-based approach for measuring, positively
impacting and tracking work performance – or work-style
transformation in corporate employees. The goal of the project
was to see if neurocognitive data and brain training
performance could benefit employees as well as provide
insights into the measurement of neurocognitive performance
in the workplace. To support the project, a large multinational
information technology equipment and services company
agreed to participate in the study. At the corporate level senior
executives were searching for methods to improve work
satisfaction and have been active in promoting work-
transformation with the goals of decreasing worker stress,
*Research supported by Fujitsu Laboratories of America, Inc. and The
Platypus Institute.
Steven L. Miller, Ph.D., Suhas Chelian, Will McBurnett and Amy A.
Kruse are with The Platypus Institute, San Diego, CA 92130 USA (phone:
+1 650-808-5539; e-mail: Steve@PlatypusNeuro.com).
extending career longevity and increasing overall happiness.
The corporation, through their own research, had become
convinced that neuroscience will play a large role in the future
of work and the implementation of artificial intelligence and
other human-machine systems. The efforts described in the
following paper were subsequently designed in full
collaboration with the corporation and the corporation waived
any rights to see individual performance data.
II. DATA COLLECTION AND PROCESSING
A. Research Design
This study used a two group, quasi-experimental, pre- and
post-test intervention design. Subjects were administered a
pre-test evaluation to establish a baseline and assigned to a
brain training program prior to group assignment. Following
6 weeks of training the subjects were re-evaluated (post-test).
A median split created two groups, based on the number of
training hours completed to compare the impact of brain
training on the independent measures of cognitive test
performance. Analysis of the pre- and post-test
electrophysiological and behavioral test scores was performed
using multivariate analysis of variances procedures.
B. Brain Training and Assessments
Subjects: Twenty-two employee volunteers were
recruited1 for the study (11 males; 11 females; average age
42.3 years). Subjects were between the ages of 18 and 65 years
and did not present co-occurring neurological disorders. The
goal of the study was to demonstrate the capability for the
deployment of online brain training to build employee
cognitive capacity.
Brain Training Program: Subjects were given the goal of
completing twenty 30-minute training sessions over a 6-week
period; for a total goal of 10 training hours. Brain training was
available on-line via computer, cellphone, etc. using a
BrainHQ.com account set up for the study. Data tracking and
program compliance updates were done weekly. The
following training areas were targeted for training using
BrainHQ: brain speed, attention, people skills and intelligence
(memory and navigation were not selected for training).
Following the completion of brain training, a median split of
the group, based on the total brain training hours, was
performed to compare the impact of brain training on a battery
of independent measures of cognitive performance.
Assessments: Following informed consent, prior to and
Winnie Tsou is with Fujitsu Laboratories of America, Inc. Sunnyvale, CA
90265 USA (phone: +1 408-530-4500; e-mail: wtsou@us.fujitsu.com)
An Investigation of Computer-based Brain Training on the
Cognitive and EEG Performance of Employees *
Steven L. Miller, Suhas Chelian, Will McBurnett, Winnie Tsou and Amy A. Kruse, Member, IEEE
following brain training, subjects performed the following
assessments with behavioral and electrophysiological data
recording: Baseline Task of Eyes Open/Eyes Closed, the
Eriksen flanker task, the DANA Standard Assessment, and
surveys on sleep, stress and emotional resilience. The DANA
Standard battery tasks were chosen because they are FDA
cleared, and easy to administer via a tablet; in addition, it is
well-normed for comparison with other groups. The present
discussion will focus on the data from the DANA tasks listed
in Table 1. For DANA tasks, a cognitive efficiency measure is
calculated which is a normalized metric (combining speed and
accuracy) of the number of correct responses per minute, see
[4] and [5]. Cognitive efficiency is used to quantify an
individual’s capacity to make correct responses per minute.
Higher scores indicate better performance.
TABLE 1: DANA STANDARD TASKS Cognitive and psychological
tasks designed to provide a standardized measure of cognitive efficiency
across repeated measures. Cognitive efficiency refers to a behavioral measure
which combines accuracy and speed; see [4] and [5].
Test Name
Task Description
Simple Reaction Time
(SRT1)
Recognize the presence of an object and tap
the object.
Procedural Reaction
Time (PRT)
Recognize 1 of 4 numbers and tap 1 of 2
buttons.
Go/No-Go Task
(GNG)
Recognize a green or gray object and only tap
in response to gray.
Code Substitution
Learning (CSL)
Recognize whether a symbol-digit pair
matches the key code shown and tap “Yes” or
“No”.
Spatial Processing
(SP)
Recognize rotation of a visual object and tap
“same” or “different”.
Matching to Sample
(MTS)
Recall a 4x4 checkboard pattern after it
disappears for 5 seconds and two options
appear.
Memory Search (MS)
Recognize letters that have been previously
memorized.
Simple Reaction Time
(SRT2)
Recognize the presence of an object and tap
the object (after ~15 minutes of cognitive
exertion).
a. For more information see Anthro tronix. com
C. EEG Data Collection
EEG data was collected with Cognionics Quick-20
headsets (Cognionics, San Diego, CA); consisting of 20 dry
electrodes with a sampling rate of 500 Hz. As shown in Fig.
1, EEG processing included the rejection of bad channels,
artifacts such as eye blinks, and separate low and high
frequency bandpass filters using NeuroPype software
(Intheon, San Diego, CA); see [6] for a representative
publication explaining the processing stages for EEG.
Power spectrum density (PSD) estimation was performed
with the Welch method and 1/f normalized. For each trial,
band power was calculated for various frequency bands and
workload was calculated using the following formula from [7]:
beta / (theta + alpha). A robust winsorized mean was used to
average across trials.
EEG was recorded during all assessments except the
surveys. Assessments took approximately 90 minutes.
For space purposes, we report the behavioral data for all
eight of the DANA Standard Battery tasks and the
electrophysiological data for two of the eight DANA tasks:
Simple Reaction Time (SRT1) and Go/No-Go task (GNG).
III. RESULTS
A. Brain Training Performance
Participants received access to a BrainHQ account
immediately following the assessment and were instructed to
complete the specified programs at least 3 times per week. The
Longer Training group had 10 participants (6 females, 4
males) in the study with an average age of 43.8 years and an
average of 30 hours trained. The Shorter Training Group had
11 subjects (5 females, 6 males) with an average age of 40.9
years and an average of 7 hours trained. During training, the
Long Training Group completed 824 levels of training
progression (i.e., higher task difficulty) compared to 201
levels completed by the Short Training Group
B. Cognitive Efficiency Results (Pre- and Post-training)
Table 2 shows the means and standard errors for the
individual behavioral tasks for each group (Long Training
Group and Short Training Group) and Time (Time 1 or before
brain training and Time 2 or after brain training). Analyses of
these pre-brain training data (Table 2, Time 1 Means) did not
reach statistical significance for any of the individual tasks. In
addition, the sum of the cognitive efficiency scores across all
the tasks for the Long and Short Training Groups was 716.2
and 715.9, respectively. Again, these differences prior to brain
training did not reach statistical significance.
After brain training, both groups showed significant
improvements on the measures of cognitive efficiency for
several DANA Standard assessments, see Table 2. The sum
across the tasks reflected an overall cognitive efficiency score
of 801.3 and 752.6 for the Long and Short Training Groups,
respectively. Higher cognitive efficiency scores reflect faster,
more accurate task performance. In this study we observed a
12% improvement in brain speed for the Long Training Group
FIGURE 1: EEG PROCESSING STEPS - Major steps in processing the
electrophysiological data. Preprocessing removes noise, artifacts, etc.;
spectral analysis finds signals of interest such as bandpower and ratios
thereof.
and a 5% increase for the Short Training Group. Differences
across the groups are significantly larger for the Longer
training group on the Procedural Reaction Time (PRT) task
and the Go/No-Go (GNG) task. Both tasks, PRT and GNG
require more than a fast response, as measured by the Simple
Reaction Time tasks (SRT1 and SRT2) and include an
additional cognitive control component to rapid response
selection.
TABLE 2 : GROUP, TASK AND TIME PERFORMANCE ON THE DANA
STANDARD TASKS - Significant Time effects for the specified Task
(p<.05) are marked with “#”; significant Group by Time effects for the
specified Task (p<.05) are marked with “*”.
Task
Time
Mean
S.E.M.
SRT1
#
Long Training Group
1
154.823
7.398
2
171.665
5.951
Short Training Group
1
152.527
6.940
2
164.847
5.582
CS
#
Long Training Group
1
42.548
3.237
2
51.277
3.234
Short Training Group
1
44.245
3.036
2
49.963
3.034
PRT
#, *
Long Training Group
1
102.120
4.225
2
114.085
3.855
Short Training Group
1
104.855
3.964
2
108.720
3.616
SP
#
Long Training Group
1
32.883
2.835
2
39.220
3.010
Short Training Group
1
32.683
2.660
2
36.239
2.824
GNG
#, *
Long Training Group
1
128.512
6.907
2
140.725
4.239
Short Training Group
1
127.235
6.480
2
127.254
3.976
MTS Long Training Group
1
39.623
3.969
2
38.648
3.423
Short Training Group
1
39.684
3.723
2
39.448
3.211
MS
#
Long Training Group
1
54.973
4.286
2
76.083
5.346
Short Training Group
1
54.838
4.021
2
65.805
5.015
SRT2 Long Training Group
1
160.709
6.065
2
169.560
6.491
Short Training Group
1
159.848
5.690
2
160.329
6.089
C. EEG Data Analyses
Due to the behavioral differences the EEG measure of
workload is limited in the current discussion to the tasks of
SRT1 and GNG. Like the behavioral measurements, before
brain training (pre-testing), subjects from both the Long
Training group and the Short Training group were
undifferentiated in the workload EEG measure for both the
SRT1 and GNG tasks (see Fig. 2). Both groups showed
moderate bilateral parietal activation and low central and
posterior activation at pre-testing.
After training (post-testing) for SRT1, both groups show
smaller workload measurements across the head, although the
decrease is larger for the Long Training Group. For example,
both groups show less bilateral parietal activation. This
parallels the behavioralboth groups performed the SRT1
task with greater efficiency. For the GNG task, however, the
changes for each group were different. After training the Long
Training Group showed decreases in the frontal regions while
the Short Training Group showed increases in the same
FIGURE 2: WORKLOAD EEG MEASUREMENTS FOR THE SIMPLE
REACTION TIME TASK (SRT1) AND THE GO/NO-GO TASK (GNG)
-Tasks as a function of training and group.
region, especially on the left side. Both groups did however
show an increase in central and posterior activation. It appears
that the Long Training Group was able to handle the task with
less workload. The behavioral data showed that the Long
Training Group performed the task better after training while
the opposite for the Short Training Group. Thus, changes in
behavioral data had corresponding changes in neural data.
III. CONCLUSION
Executive functions (e.g., information processing,
sequencing, decision making, planning, etc.) are associated
with optimal cognitive performance and are also known to
contribute to corporate work tasks [1]. In the present
discussion we have demonstrated that independent computer-
based brain assessment (DANA) and training (BrainHQ) could
provide a scalable solution to evaluate and develop executive
functions, functions that are malleable throughout the life-
span. BrainHQ training increased brain processing speed as
measured by the DANA Standard battery on a variety of
neurobehavioral tasks. Further, the independently developed
DANA Standard battery [5] and BrainHQ training program [8]
cross-validate their respective evaluation and training of brain
speed. The further elaboration of the neuroplastic mechanisms
that may underly these behavioral changes appear to be
clarified by an electrophysiological measure of workload. The
next stage of research in this area will include greater rigor
along several dimensions such as: more subjects, randomized
assignment to groups with an active control group, detailed
statistical analysis of EEG data, parameterization of where the
workload EEG measure is appropriate (as well as other
measures such as attention or memory), and so on. This
information will further optimize and personalize brain
training.
Overall, the corporate study demonstrated positive benefits
for the group of participants in several areas of neurocognitive
performance. Further, significantly higher gains were recorded
in the highest training group with moderate gains in the lower
training group. With additional EEG-based analysis, we will
be able to refine our understanding into the mechanisms of
neuroplasticity that occurred as a direct result of our program.
More importantly, with this study, we demonstrate that a
cognitive state (e.g., workload performance) could support the
further extension of real-time brain performance evaluations
in the corporate environment. The loop of “measure-boost-
track” was shown to be effective both qualitatively and
quantitatively and worthwhile results were seen with modest
training, gains in attention, executive control and decision-
making systems were present. Finally, while the study was not
designed to elucidate the “dose response” of cognitive training
there does appear to be some value in extending further
research in determining the dose-response curve for brain
training benefits as well as the extension of the nature of the
benefits to specific corporate tasks (e.g., bookkeeping, digital
correspondence).
APPENDIX
EEG bands were defined as follows: delta [1, 3], theta [4,
7], alpha [8, 15], beta [16, 31], gamma [32, 40]; all intervals
are closed on both ends.
ACKNOWLE DGMENT
We would like to recognize the efforts of Fujitsu
Laboratories of America, Inc. for sponsoring the study. In
addition, we would like to thank the executive team at
Intheon.io for assistance in the analysis of the
electrophysiological data and executives at Anthrotronix and
Posit Science for in-depth guidance and support on the use of
their respective products to support the study.
REFERENCES
[1] F. Schmidt, "The role of general cognitive ability and job performance:
why there cannot be a debate," Human Performance, vol. 15, pp. 187-
210, 2002.
[2] J. Edwards et al., “Speed of processing training results in lower risk of
dementia,” Alzheimer’s Dementia, vol. 3, pp. 603611, 2017.
[3] A. Lampit, H. Hallock, and M. Valenzuela, “Computerized cognitive
training in cognitively healthy older adults: a systematic review and
meta-analysis of effect modifiers,” PLoS Med., vol. 11, e1001756,
2014.
[4] I. Coffman et al., “Computerized cognitive testing norms in active-duty
military personnel: potential for contamination by psychologically
unhealthy individuals,” Applied Neuropsychology: Adult, vol. 25, pp.
497-503, 2018.
[5] AnthroTronix, DANA: the brain thermometer, user guide v6 for iOS.
Silver Spring, MD: Author, 2018.
[6] N. Bigdely-Shamlo et al., “The PREP pipeline: standardized
preprocessing for large-scale EEG analysis.” Frontiers in
Neuroinformatics, vol. 9, pp. 16, 2015.
[7] L. Prinzel et al., "A closed-loop system for examining
psychophysiological measures for adaptive task allocation," The
International Journal of Aviation Psychology, vol. 10, pp. 393-410,
2000.
[8] BrainHQ from Posit Science. “BrainHQ from Posit Science. Internet:
https://www.brainhq.com/ [Accessed: Feb. 02, 2019].
Notes: 1 Subjects were recruited for the study after attending a research
briefing by two members of the Platypus Institute research team. The
participating organizations provided written assurances that all reasonable
efforts would be made to keep their individual performance and EEG data
anonymous. Subjects provided informed consent prior to the start of the
study and could terminate their participation without consequence. Subjects
received modest compensation for their participation.
ResearchGate has not been able to resolve any citations for this publication.
  • DANA: the brain thermometer, user guide v6 for iOS
    • Anthrotronix
    AnthroTronix, DANA: the brain thermometer, user guide v6 for iOS. Silver Spring, MD: Author, 2018.
  • A closed-loop system for examining psychophysiological measures for adaptive task allocation
    L. Prinzel et al., "A closed-loop system for examining psychophysiological measures for adaptive task allocation," The International Journal of Aviation Psychology, vol. 10, pp. 393-410, 2000.
  • BrainHQ from Posit Science
    BrainHQ from Posit Science. "BrainHQ from Posit Science. Internet: https://www.brainhq.com/ [Accessed: Feb. 02, 2019].
  • AnthroTronix, DANA: the brain thermometer, user guide v6 for iOS
    • Anthrotronix
  • Article
    Full-text available
    Introduction Cognitive training improves cognitive performance and delays functional impairment, but its effects on dementia are not known. We examined whether three different types of cognitive training lowered the risk of dementia across 10 years of follow-up relative to control and if greater number of training sessions attended was associated with lower dementia risk. Methods The Advanced Cognitive Training in Vital Elderly (NCT00298558) study was a randomized controlled trial (N = 2802) among initially healthy older adults, which examined the efficacy of three cognitive training programs (memory, reasoning, or speed of processing) relative to a no-contact control condition. Up to 10 training sessions were delivered over 6 weeks with up to four sessions of booster training delivered at 11 months and a second set of up to four booster sessions at 35 months. Outcome assessments were taken immediately after intervention and at intervals over 10 years. Dementia was defined using a combination of interview- and performance-based methods. Results A total of 260 cases of dementia were identified during the follow-up. Speed training resulted in reduced risk of dementia (hazard ratio [HR] 0.71, 95% confidence interval [CI] 0.50–0.998, P = .049) compared to control, but memory and reasoning training did not (HR 0.79, 95% CI 0.57–1.11, P = .177 and HR 0.79, 95% CI 0.56–1.10, P = .163, respectively). Each additional speed training session was associated with a 10% lower hazard for dementia (unadjusted HR, 0.90; 95% CI, 0.85–0.95, P < .001). Discussion Initially, healthy older adults randomized to speed of processing cognitive training had a 29% reduction in their risk of dementia after 10 years of follow-up compared to the untreated control group.
  • Article
    Normative reference data used for clinical interpretation of neuropsychological testing results are only valid to the extent that the sample they are based on is composed of “normal” individuals. Accordingly, efforts are made to exclude individuals with histories and/or diagnoses that might bias test performance. In this report, we focus on these features in active-duty military personnel because published data on computerized neurocognitive testing norms for this population have not explicitly considered the consequences of neurobehavioral disorders (e.g., PTSD, depression), which are prevalent in this population and known to affect performance on some cognitive assessments. We administered DANA, a mobile, neurocognitive assessment tool, to a large sample of active-duty military personnel and found that scores on self-administered psychological assessments negatively impacted a number of neurocognitive tests. These results suggest that neurobehavioral disorders that are relatively common in this population should be controlled for when establishing normative datasets for neurocognitive outcomes.
  • Article
    Full-text available
    The technology to collect brain imaging and physiological measures has become portable and ubiquitous, opening the possibility of large-scale analysis of real-world human imaging. By its nature, such data is large and complex, making automated processing essential. This paper shows how lack of attention to the very early stages of an EEG preprocessing pipeline can reduce the signal-to-noise ratio and introduce unwanted artifacts into the data, particularly for computations done in single precision. We demonstrate that ordinary average referencing improves the signal-to-noise ratio, but that noisy channels can contaminate the results. We also show that identification of noisy channels depends on the reference and examine the complex interaction of filtering, noisy channel identification, and referencing. We introduce a multi-stage robust referencing scheme to deal with the noisy channel-reference interaction. We propose a standardized early-stage EEG processing pipeline (PREP) and discuss the application of the pipeline to more than 600 EEG datasets. The pipeline includes an automatically generated report for each dataset processed. Users can download the PREP pipeline as a freely available MATLAB library from http://eegstudy.org/prepcode.
  • Article
    Full-text available
    Background: New effective interventions to attenuate age-related cognitive decline are a global priority. Computerized cognitive training (CCT) is believed to be safe and can be inexpensive, but neither its efficacy in enhancing cognitive performance in healthy older adults nor the impact of design factors on such efficacy has been systematically analyzed. Our aim therefore was to quantitatively assess whether CCT programs can enhance cognition in healthy older adults, discriminate responsive from nonresponsive cognitive domains, and identify the most salient design factors. Methods and findings: We systematically searched Medline, Embase, and PsycINFO for relevant studies from the databases' inception to 9 July 2014. Eligible studies were randomized controlled trials investigating the effects of ≥ 4 h of CCT on performance in neuropsychological tests in older adults without dementia or other cognitive impairment. Fifty-two studies encompassing 4,885 participants were eligible. Intervention designs varied considerably, but after removal of one outlier, heterogeneity across studies was small (I(2) = 29.92%). There was no systematic evidence of publication bias. The overall effect size (Hedges' g, random effects model) for CCT versus control was small and statistically significant, g = 0.22 (95% CI 0.15 to 0.29). Small to moderate effect sizes were found for nonverbal memory, g = 0.24 (95% CI 0.09 to 0.38); verbal memory, g = 0.08 (95% CI 0.01 to 0.15); working memory (WM), g = 0.22 (95% CI 0.09 to 0.35); processing speed, g = 0.31 (95% CI 0.11 to 0.50); and visuospatial skills, g = 0.30 (95% CI 0.07 to 0.54). No significant effects were found for executive functions and attention. Moderator analyses revealed that home-based administration was ineffective compared to group-based training, and that more than three training sessions per week was ineffective versus three or fewer. There was no evidence for the effectiveness of WM training, and only weak evidence for sessions less than 30 min. These results are limited to healthy older adults, and do not address the durability of training effects. Conclusions: CCT is modestly effective at improving cognitive performance in healthy older adults, but efficacy varies across cognitive domains and is largely determined by design choices. Unsupervised at-home training and training more than three times per week are specifically ineffective. Further research is required to enhance efficacy of the intervention. Please see later in the article for the Editors' Summary.
  • Article
    Full-text available
    Given the overwhelming research evidence showing the strong link between general cognitive ability (GCA) and job performance, it is not logically possible for industrial -organizational (I/O) psychologists to have a serious debate over whether GCA is important for job performance. However, even if none of this evidence existed in I/O psychology, research findings in differential psychology on the nature and correlates of GCA provide a sufficient basis for the conclusion that GCA is strongly related to job performance. In I/O psychology, the theoretical basis for the empirical evidence linking GCA and job performance is rarely presented, but is critical to understanding and acceptance of these findings. The theory explains the why behind the empirical findings. From the viewpoint of the kind of world we would like to live in - and would like to believe we live in - the research findings on GCA are not what most people would hope for and are not welcome. However, if we want to remain a science-based field, we cannot reject what we know to be true in favor of what we would like to be true.