High Cognitive Load Research Articles

Cognitive load scale in learning formal definition of limit: A rasch model approach

Constructing proofs for the limit using the formal definition induces a high cognitive load. Common assessment tools, like cognitive load scales, lack specificity for the concept of limits. This research aims to validate an instrument tailored to assess cognitive load in students focused on the formal definition of limits, addressing the need for diverse strategies in education. The research employs a quantitative survey design with a Rasch model approach, utilizing a data collection instrument in the form of a questionnaire. Subsequently, the data are analyzed by focusing on three aspects: (1) item fit to the Rasch model, (2) unidimensionality, and (3) rating scale. A total of 315 students from three private universities in Banten participated as research respondents. The findings of this study affirm the validity of the cognitive load scale centered on the formal definition of limit, meeting the stringent standards set by Rasch modeling. Additionally, the results of the study provide evidence of the scale’s adherence to the monotonic principle of the Rasch model. These outcomes contribute to a comprehensive understanding of cognitive load in the context of learning formal definition of limit, providing a solid foundation for instructional design and assessment strategies.

The Impact of Patient Race, Patient Socioeconomic Status, and Cognitive Load of Physician Residents and Fellows on Chronic Pain Care Decisions

Social stereotypes are more likely to influence decision-making under conditions of high cognitive load (ie, mental workload), such as in medical settings. We examined how patient race, patient socioeconomic status (SES), physician cognitive load, and physician implicit beliefs about race and SES differences in pain tolerance impacted physicians’ pain treatment decisions. Physician residents and fellows (N = 120) made treatment decisions for 12 computer-simulated patients with back pain that varied by race (Black/White) and SES (low/high). Half of the physicians were randomized to be interrupted during the decision task to make hypertension medication conversion calculations (high cognitive load group), while the other half completed the task without interruptions (low cognitive load group). Both groups were given equal time to make pain care decisions (2.5 minutes/patient). Results of multilevel ordinal logistic regression analyses indicated that physicians prescribed weaker analgesics to patients with high vs. low SES (odds ratio = .68, 95% confidence interval [.48, .97], P = .03). There was also a patient SES-by-cognitive load interaction (odds ratio = .56, 95% confidence interval [.31, 1.01], P = .05) that is theoretically and potentially practically meaningful but was not statistically significant at P < .05. These findings shed light on physician cognitive load as a clinically-relevant factor in the context of pain care quality and equity. PerspectiveThese findings highlight the clinical relevance of physician cognitive load (eg, mental workload) when providing pain care for diverse patients. This line of work can support the development of interventions to manage physician cognitive load and its impact on pain care, which may ultimately help reduce pain disparities.

Collaborative robots can augment human cognition in regret-sensitive tasks.

Despite theoretical benefits of collaborative robots, disappointing outcomes are well documented by clinical studies, spanning rehabilitation, prostheses, and surgery. Cognitive load theory provides a possible explanation for why humans in the real world are not realizing the benefits of collaborative robots: high cognitive loads may be impeding human performance. Measuring cognitive availability using an electrocardiogram, we ask 25 participants to complete a virtual-reality task alongside an invisible agent that determines optimal performance by iteratively updating the Bellman equation. Three robots assist by providing environmental information relevant to task performance. By enabling the robots to act more autonomously-managing more of their own behavior with fewer instructions from the human-here we show that robots can augment participants' cognitive availability and decision-making. The way in which robots describe and achieve their objective can improve the human's cognitive ability to reason about the task and contribute to human-robot collaboration outcomes. Augmenting human cognition provides a path to improve the efficacy of collaborative robots. By demonstrating how robots can improve human cognition, this work paves the way for improving the cognitive capabilities of first responders, manufacturing workers, surgeons, and other future users of collaborative autonomy systems.

Unraveling the Complexities of Second Language Lexical Stress Processing: The Impact of First Language Transfer, Second Language Proficiency, and Exposure

AbstractWe investigated the role of cue weighting, second language (L2) proficiency, and L2 daily exposure in L2 learning of suprasegmentals different from the first language (L1), using eye‐tracking. Spanish monolinguals, English–Spanish learners, and Mandarin–Spanish learners saw a paroxytone and an oxytone verb (e.g., FIRma–firMÓ “s/he signs–signed”), listened to a sentence containing one of the verbs, and chose the one that they heard. The three languages have contrastive lexical stress, but suprasegmentals have a greater functional load in Mandarin than in English. Monolinguals predicted suffixes accurately with both stress conditions and favored oxytones, but learners predicted suffixes accurately only with oxytones, the condition activating fewer lexical competitors. Monolinguals predicted suffixes accurately sooner but at a slower rate than did learners. L2 proficiency, but not L1 or L2 exposure, facilitated L2 predictions. In conclusion, learners of a contrastive‐stress L1 rely on L2 suprasegmentals to the same extent as monolinguals, regardless of their L1. Lower L2 proficiency and higher cognitive load (more lexical competitors) reduce learners’ reliance on suprasegmentals.

Storage interference in working memory cannot be removed by attention

In the present study, we examined the hypothesis that the storage interference in working memory can be removed by attention. A dual-task paradigm was employed in Experiment 1 and 2, in which participants performed a color memory task and an RSVP letter detection task concurrently. The cognitive load of the RSVP letter detection task and the storage interference caused by the RSVP letter detection task was manipulated independently. That is, the produced storage-interference difference between the low and high interference conditions was comparable between the low and high cognitive load conditions, whereas the available attentional resources were different under the two cognitive load conditions. Since there were more attentional resources in the low load condition, the removal hypothesis predicts that differences in recall performance between the low and high interference conditions should be larger in high load than in low load, i.e., there would be an interaction between load and interference. However, the results of the two experiments did not show such an interaction. In Experiment 3, we manipulated the time available for the removal mechanism to work while inducing both the storage interference and processing interference. The results showed no sign of interference removal. Thus, the present results provided solid evidence to challenge the removal hypothesis.

Intraoperative Cognitive Load Differences between Trainees and Attending Surgeons

Background/Objective: Cognitive load (CL) is the amount of mental effort and resources required to complete a task and process information in the working memory. In a surgical setting, high CL decreases attention to critical details, and slows down decision-making, increasing errors which can compromise patient safety. Our aim was to determine if CL differences exist between trainees and experienced attending surgeons. Such differences could inform readiness for trainee autonomy and training paradigms.&#x0D; Methods: Eye trackers were used by attending general surgeons and trainees during robotic gastrointestinal procedures to determine their CL. Average fixation rate (AFR), fixation to saccade (F:S) ratio, and change in pupil size, were recorded and compared between attending surgeons and trainees to determine differences in intraoperative cognitive load.&#x0D; Results: Three attendings and three trainees participated. While operating, trainees had a lower AFR (M=0.775, SD= 0.093) and a lower F:S (M=0.497, SD=0.102) compared to attendings (M=0.842, SD=0.152 and M=0.592, SD=0.243, respectively). Both trainees and attendings had a larger pupil size (M=4.29, SD=0.773 and M=3.63, SD=0.077, respectively) while operating compared to their pupil size while not operating (M=4.057, SD=0.821 and M=3.496, SD=0.059, respectively.&#x0D; Conclusion and Potential Impact:In this pilot study, we determined eye tracking metrics can be used to detect differences in intraoperative CL between trainees and attending surgeons and while they were operating or observing. Further research should determine methods that decrease CL of trainees.

Quality and quantity: How contexts influence the emergence of teacher bias

A growing body of research has demonstrated that teachers’ judgements may be biased by the demographics and characteristics of the students they teach. However, less work has investigated the contexts in which teachers may be most vulnerable to bias. In two pre-registered experimental studies we explored whether the quality of students’ work, and the cognitive load placed on the grader, would influence the emergence of biases relating to socioeconomic status (SES) and ethnicity. In Study 1, teachers (N = 397) graded work of either high or low quality that had ostensibly been written by a student who varied in terms of SES and ethnicity. We found that SES and—to a lesser extent—ethnicity biases were more likely to manifest when the standard of work was below average, largely favouring a student from an affluent White background. In Study 2, an undergraduate sample (N = 334) provided judgements based on an identical piece of work written by a student who varied by SES and ethnicity. Importantly, they formulated these judgements whilst working under high or low cognitive load, which was manipulated via a simultaneous listening task. Results showed that under high cognitive load, a student from an affluent White British background was afforded a boost in grading that was not afforded to Black Caribbean and/or lower SES students. These findings highlight contexts in which teachers may be most prone to biased judgements and should be used by educational institutions to de-bias their workflows, workloads and workforces.

How time flies: Time perception and intertemporal choice

This paper identifies a potential gap between time preference and intertemporal choices such as time discounting tasks: The intertemporal choices could be partly driven by a biased perception of time and thus may not completely reveal actual time preference. To test this, we explore the causal relationship between time perception and intertemporal choices by conducting a laboratory experiment in which cognitive load is used as a stimulating instrument to induce differences in time perception. We establish that the perceived time lengths for subjects with high cognitive load are shorter than those with low cognitive load and that individuals who underestimate time appear more patient in their intertemporal choices. Mediation analyses show that time perception mediates a significant portion of cognitive load’s effect on intertemporal choices. Our study thus demonstrates that time preference identified by intertemporal choices might be confounded by potentially biased time perception, calling for improving suboptimal time-related economic decisions due to an individual’s misperception of how time flies.

Extraneous Load, Patient Census, and Patient Acuity Correlate With Cognitive Load During ICU Rounds

BackgroundCognitive Load Theory asserts that learning and performance degrade when cognitive load exceeds working memory capacity. This is particularly relevant in the learning environment of intensive care unit (ICU) rounds, when multidisciplinary providers integrate complex decision-making and teaching in a noisy, high-stress environment prone to cognitive distractions. Research QuestionWhat features of ICU rounds correlate with high provider cognitive load? Study Design and MethodsThis was an observational, multisite study of multidisciplinary providers during ICU rounds. Investigators recorded rounding characteristics and hourly extraneous cognitive load events during rounds (defined as distractions, episodes of split-attention or repetition, and deviations from standard communication format). Following rounds, investigators measured each provider’s cognitive load using the Provider Task Load (PTL), an instrument derived from the NASA Task Load Index survey that assesses perceived workload associated with complex tasks. Relationships between rounding characteristics, extraneous load, and PTL score were evaluated using mixed-effects modeling. Results76 providers were observed during 32 rounds from December 2020 – May 2021. The mean (±standard deviation) rounding census was 12.5±2.9 patients. The mean rounding time was 2h17m±49m. The mean extraneous load was 20.5±4.5 events per hour, or one event every 2m51sec. This included 8.6±3.4 distractions, 8.2±4.2 communication deviations, 1.9±1.4 repetitions, and 1.8±1.3 episodes of split-attention per hour. Controlling for covariates, the hourly extraneous load events, number of new patients, and number of higher acuity patients were each associated with increased PTL score (slope: 2.40 and 95% CI: 0.76 – 4.04, slope: 5.23 and 95% CI: 2.02 – 8.43, slope: 3.35 and 95% CI: 1.34 – 5.35, respectively). InterpretationIncreased extraneous load, new patients, and patient acuity were associated with higher cognitive load during ICU rounds. These results can help direct how the ICU rounding structure may be modified to reduce workload and optimize provider learning and performance.

The impact of perceptual complexity on road crossing decisions in younger and older adults

Cognitive abilities decline with healthy ageing which can have a critical impact on day-to-day activities. One example is road crossing where older adults (OAs) disproportionally fall victim to pedestrian accidents. The current research examined two virtual reality experiments that investigated how the complexity of the road crossing situation impacts OAs (N = 19, ages 65–85) and younger adults (YAs, N = 34, ages 18–24) with a range of executive functioning abilities (EFs). Overall, we found that OAs were able to make safe crossing decisions, and were more cautious than YAs. This continued to be the case in high cognitive load situations. In these situations, safe decisions were associated with an increase in head movements for participants with poorer attention switching than participants with better attention switching suggesting these groups developed compensation strategies to continue to make safe decisions. In situations where participants had less time to make a crossing decision all participants had difficulties making safe crossing decisions which was amplified for OAs and participants with poorer EFs. Our findings suggest more effort should be taken to ensure that road crossing points are clear of visual obstructions and more speed limits should be placed around retirement or care homes, neither of which are legislated for in the UK and Australia.

The Right-Ear Advantage in Static and Dynamic Cocktail-Party Situations.

When presenting two competing speech stimuli, one to each ear, a right-ear advantage (REA) can often be observed, reflected in better speech recognition compared to the left ear. Considering the left-hemispheric dominance for language, the REA has been explained by superior contralateral pathways (structural models) and language-induced shifts of attention to the right (attentional models). There is some evidence that the REA becomes more pronounced, as cognitive load increases. Hence, it is interesting to investigate the REA in static (constant target talker) and dynamic (target changing pseudo-randomly) cocktail-party situations, as the latter is associated with a higher cognitive load than the former. Furthermore, previous research suggests an increasing REA, when listening becomes more perceptually challenging. The present study examined the REA by using virtual acoustics to simulate static and dynamic cocktail-party situations, with three spatially separated talkers uttering concurrent matrix sentences. Sentences were presented at low sound pressure levels or processed with a noise vocoder to increase perceptual load. Sixteen young normal-hearing adults participated in the study. The REA was assessed by means of word recognition scores and a detailed error analysis. Word recognition revealed a greater REA for the dynamic than for the static situations, compatible with the view that an increase in cognitive load results in a heightened REA. Also, the REA depended on the type of perceptual load, as indicated by a higher REA associated with vocoded compared to low-level stimuli. The results of the error analysis support both structural and attentional models of the REA.

Cognitive and perceptual load have opposing effects on brain network efficiency and behavioral variability in ADHD.

Attention-deficit/hyperactivity disorder (ADHD) is a highly prevalent neurodevelopmental disorder associated with suboptimal outcomes throughout the life-span. Extant work suggests that ADHD-related deficits in task performance may be magnified under high cognitive load and minimized under high perceptual load, but these effects have yet to be systematically examined, and the neural mechanisms that undergird these effects are as yet unknown. Herein, we report results from three experiments investigating how performance in ADHD is modulated by cognitive load and perceptual load during a naturalistic task. Results indicate that cognitive load and perceptual load influence task performance, reaction time variability (RTV), and brain network topology in an ADHD-specific fashion. Increasing cognitive load resulted in reduced performance, greater RTV, and reduced brain network efficiency in individuals with ADHD relative to those without. In contrast, increased perceptual load led to relatively greater performance, reduced RTV, and greater brain network efficiency in ADHD. These results provide converging evidence that brain network efficiency and intraindividual variability in ADHD are modulated by both cognitive and perceptual load during naturalistic task performance.

Evaluation of Parkinson's disease early diagnosis using single-channel EEG features and auditory cognitive assessment.

Parkinson's disease (PD) often presents with subtle early signs, making diagnosis difficult. F-DOPA PET imaging provides a reliable measure of dopaminergic function and is a primary tool for early PD diagnosis. This study aims to evaluate the ability of machine-learning (ML) extracted EEG features to predict F-DOPA results and distinguish between PD and non-PD patients. These features, extracted using a single-channel EEG during an auditory cognitive assessment, include EEG feature A0 associated with cognitive load in healthy subjects, and EEG feature L1 associated with cognitive task differentiation. Participants in this study are comprised of cognitively healthy patients who had undergone an F-DOPA PET scan as a part of their standard care (n = 32), and cognitively healthy controls (n = 20). EEG data collected using the Neurosteer system during an auditory cognitive task, was decomposed using wavelet-packet analysis and machine learning methods for feature extraction. These features were used in a connectivity analysis that was applied in a similar manner to fMRI connectivity. A preliminary model that relies on the features and their connectivity was used to predict initially unrevealed F-DOPA test results. Then, generalized linear mixed models (LMM) were used to discern between PD and non-PD subjects based on EEG variables. The prediction model correctly classified patients with unrevealed scores as positive F-DOPA. EEG feature A0 and the Delta band revealed distinct activity patterns separating between study groups, with controls displaying higher activity than PD patients. In controls, EEG feature L1 showed variations between resting state and high-cognitive load, an effect lacking in PD patients. Our findings exhibit the potential of single-channel EEG technology in combination with an auditory cognitive assessment to distinguish positive from negative F-DOPA PET scores. This approach shows promise for early PD diagnosis. Additional studies are needed to further verify the utility of this tool as a potential biomarker for PD.

Distraction suppresses high-fat flavor perception

Distraction during eating contributes to overeating, and when habitually eating with distraction, this may contribute to the development of obesity. One of the proposed mediating mechanisms is the suppression of intensity perception in odor and taste. The effect of distraction on fat intensity perception in flavor, the multisensory combination of odor, taste, and other sensory aspects, is still unknown.In this study, 32 participants (22 women) performed a flavor perception task while also performing a distracting working memory task. In each trial, participants were instructed to observe and memorize a string of 3 (low cognitive load) or 7 (high cognitive load) consonants. Then they received a small quantity of a high- or low-fat chocolate drink, and after that, they were asked to select the string they tried to memorize from three answer options. Last, they rated the intensity and fattiness of the flavor.As intended, in the working memory task, we observed that with a high cognitive load (relative to a low cognitive load), accuracy decreased and response times increased. Regarding perception of the flavors, we observed that overall, high-fat drinks were rated as more intense and fattier. Cognitive load and fat content interacted, such that for the low-fat drink, intensity and fattiness ratings were similar under both cognitive loads; however, under the high cognitive load (relative to the low cognitive load), intensity and fattiness ratings for the high-fat drink were lower.Our results show that distraction can impact the perception of fat in high-fat drinks. If distraction primarily reduces perception of high-fat foods, this may pose a particular risk of overeating high-calorie foods.

Distinct age-related brain activity patterns in the prefrontal cortex when increasing cognitive load: A functional near-infrared spectroscopy study

Researchers have long observed distinct brain activity patterns in older adults compared with younger adults that correlate with cognitive performance. Mainly, older adults tend to show over-recruitment of bilateral brain regions during lower task loads and improved performance interpreted as compensation, but not observed at higher loads. However, there are discrepancies about whether increases in activity are compensatory and whether older adults can show compensation at higher loads. Our aim was to examine age-related differences in prefrontal cortex (PFC) activity and cognitive performance using functional near-infrared spectroscopy (fNIRS) during single and dual N-back tasks. Twenty-seven young adults (18–27 years) and 31 older adults (64–84 yrs) took part in the study. We used a robust fNIRS data methodology consisting of channel and region of interest analyses. Results showed differences in performance between task load conditions and age-related differences in reaction times but no age-group effects for accuracy. Older adults exhibited more bilateral PFC activation compared with young adults across all tasks and showed increases in brain activity in high compared to low load conditions. Our findings further support previous reports showing that older adults use compensatory recruitment of additional brain regions in PFC to maintain cognitive performance but go against the notion that such compensation is not present at higher cognitive loads. Additionally, our results indicate that fNIRS is a sensitive tool that can characterize adaptive cortical changes in healthy aging.

Validation of a mobile fNIRS device for measuring working memory load in the prefrontal cortex

Functional near-infrared spectroscopy (fNIRS) is a neuroimaging technique that measures cortical blood flow to infer neural activation. Traditionally limited to laboratory settings due to high costs and complex operation, recent advancements have introduced mobile fNIRS devices, significantly broadening the scope of potential research participants. This study validates the use of the Mendi, a two-channel mobile fNIRS system, for measuring prefrontal oxyhemoglobin concentration changes during an n-back task. We manipulated task difficulty through different n-back levels (one-back versus three-back), revealing increased oxyhemoglobin concentrations in the prefrontal cortex during the more demanding three-back task compared to the one-back task. This finding demonstrates the Mendi's ability to distinguish between low and high cognitive task loads. Behavioural data, showing a decrease in accuracy under high load conditions, further corroborates these neuroimaging findings. Our study validates the Mendi mobile fNIRS system as an effective tool for assessing working memory load and underscores its potential in enhancing neuroscientific research accessibility. The user-friendly and cost-effective nature of mobile fNIRS systems like the Mendi opens up neuroscientific research to a diverse set of participants, enabling the investigation of neural processes in real-world environments across a variety of demographic groups.

Event Related Brain Responses Reveal the Impact of Spatial Augmented Reality Predictive Cues on Mental Effort.

This article presents the results from a Spatial Augmented Reality (SAR) study which evaluated the cognitive cost of several predictive cues. Participants performed a validated procedural button pressing task, where the predictive cue annotations guided them to the upcoming task. While existing research has evaluated predictive cues based on their performance and self-rated mental effort, actual cognitive cost has yet to be investigated. To measure the user's brain activity, this study utilized electroencephalogram (EEG) recordings. Cognitive load was evaluated by measuring brain responses for a secondary auditory oddball task, with reduced brain responses to oddball tones expected when cognitive load in the primary task is highest. A simple monitor n-back task and procedural task comparing monitor versus SAR were conducted, followed by a version of the procedural task comparing the SAR predictive cues. Results from the brain responses were able to distinguish between performance enhancing cues with a high and low cognitive load. Electrical brain responses also revealed that having an arc or arrow guide towards the upcoming task required the least amount of mental effort.

Continuous Physiological Monitoring of the Combined Exposure to Hypoxia and High Cognitive Load in Military Personnel.

Military aviators endure high cognitive loads and hypoxic environments during flight operations, impacting the autonomic nervous system (ANS). The synergistic effects of these exposures on the ANS, however, are less clear. This study investigated the simultaneous effects of mild hypoxia and high cognitive load on the ANS in military personnel. This study employed a two-factor experimental design. Twenty-four healthy participants aged between 19 and 45 years were exposed to mild hypoxia (14.0% O2), normoxia (21.0% O2), and hyperoxia (33.0% O2). During each epoch (n = 5), participants continuously performed one 15 min and one 10 min series of simulated, in-flight tasks separated by 1 min of rest. Exposure sequences (hypoxia-normoxia and normoxia-hyperoxia) were separated by a 60 min break. Heart rate (HR), heart rate variability (HRV), and O2 saturation (SpO2) were continuously measured via an armband monitor (Warfighter MonitorTM, Tiger Tech Solutions, Inc., Miami, FL, USA). Paired and independent t-tests were used to evaluate differences in HR, HRV, and SpO2 within and between exposure sequences. Survival analyses were performed to assess the timing and magnitude of the ANS responses. Sympathetic nervous system (SNS) activity during hypoxia was highest in epoch 1 (HR: +6.9 bpm, p = 0.002; rMSSD: -9.7 ms, p = 0.003; SDNN: -11.3 ms, p = 0.003; SpO2: -8.4%, p < 0.0000) and appeared to slightly decline with non-significant increases in HRV. During normoxia, SNS activity was heightened, albeit non-significantly, in epoch 1, with higher HR (68.5 bpm vs. 73.0 bpm, p = 0.06), lower HRV (rMSSD: 45.1 ms vs. 38.7 ms, p = 0.09 and SDNN: 52.5 ms vs. 45.1 ms, p = 0.08), and lower SpO2 (-0.7% p = 0.05). In epochs 2-4, HR, HRV, and SpO2 trended towards baseline values. Significant between-group differences in HR, HRV, and O2 saturation were observed. Hypoxia elicited significantly greater HRs (+5.0, p = 0.03), lower rMSSD (-7.1, p = 0.03), lower SDNN (-8.2, p = 0.03), and lower SpO2 (-1.4%, p = 0.002) compared to normoxia. Hyperoxia appeared to augment the parasympathetic reactivation reflected by significantly lower HR, in addition to higher HRV and O2 relative to normoxia. Hypoxia induced a greater ANS response in military personnel during the simultaneous exposure to high cognitive load. The significant and differential ANS responses to varying O2 levels and high cognitive load observed highlight the importance of continuously monitoring multiple physiological parameters during flight operations.

Physical Activity in Turkish Esports Players: Age Differences Approach

Esports, which is a sport with high cognitive load, is thought to have certain dark side points from a sustainable health perspective due to the prolonged seated position. In this study, the focus has been on addressing the dark points through the physical activity levels and daily gaming durations of Turkish esports players. The research included 373 esports players between the ages of 14-25 residing in different cities of Turkey. The participants were divided into 3 categories: Adolescents (14-16), Late Adolescents (17-19), and Adults (20-24). The participants' demographic information, International Physical Activity Questionnaire (IPAQ) data, and daily gaming durations were collected through a web-based survey program. IPAQ and gaming times were separately analyzed using one-way independent measures ANOVA to compare three different age groups (Adolescent, Late Adolescent, and Young Adult). Results showed that there was a significant effect of age on gaming years, F(2, 370) = 19.84, p &amp;lt; .001, ω2 = .09 (medium effect), high-level physical activity F(2, 370) = 3.33, p &amp;lt; .05, ω2 = .01 (small effect) and sitting time scores F(2, 370) = 6.15, p &amp;lt; .01, ω2 = .02. The study results indicate that as years of e-gaming experience increase, it reveals the risk of 'professionalization '-based sedentary behavior.

Characterisation of Cognitive Load Using Machine Learning Classifiers of Electroencephalogram Data.

A high cognitive load can overload a person, potentially resulting in catastrophic accidents. It is therefore important to ensure the level of cognitive load associated with safety-critical tasks (such as driving a vehicle) remains manageable for drivers, enabling them to respond appropriately to changes in the driving environment. Although electroencephalography (EEG) has attracted significant interest in cognitive load research, few studies have used EEG to investigate cognitive load in the context of driving. This paper presents a feasibility study on the simulation of various levels of cognitive load through designing and implementing four driving tasks. We employ machine learning-based classification techniques using EEG recordings to differentiate driving conditions. An EEG dataset containing these four driving tasks from a group of 20 participants was collected to investigate whether EEG can be used as an indicator of changes in cognitive load. The collected dataset was used to train four Deep Neural Networks and four Support Vector Machine classification models. The results showed that the best model achieved a classification accuracy of 90.37%, utilising statistical features from multiple frequency bands in 24 EEG channels. Furthermore, the Gamma and Beta bands achieved higher classification accuracy than the Alpha and Theta bands during the analysis. The outcomes of this study have the potential to enhance the Human-Machine Interface of vehicles, contributing to improved safety.