Adrian Curtin

Monitoring expertise development during simulated UAV piloting tasks using optical brain imaging

An accurate assessment of mental workload and expertise level would help improve operational safety and efficacy of human computer interaction for aerospace applications. The current study utilized functional near-infrared spectroscopy (fNIR) to investigate the relationship of the hemodynamic response in the anterior prefrontal cortex to changes in mental workload, level of expertise, and task performance during learning of simulated unmanned aerial vehicle (UAV) piloting tasks. Results indicated that fNIR measures are correlated to task performance and subjective self-reported measures; and contained additional information that allowed categorizing learning phases. Level of expertise does appear to influence the hemodynamic response in the dorsolateral/ventrolateral prefrontal cortices. Since fNIR allows development of portable and wearable instruments, it has the potential to be deployed in future learning environments to personalize the training regimen and/or assess the effort of human operators in critical multitasking settings.

Towards a Hybrid P300-Based BCI Using Simultaneous fNIR and EEG

Next generation brain computer interfaces (BCI) are expected to provide robust and continuous control mechanism. In this study, we assessed integration of optical brain imaging (fNIR: functional near infrared spectroscopy) to a P300-BCI for improving BCI usability by monitoring cognitive workload and performance. fNIR is a safe and wearable neuroimaging modality that tracks cortical hemodynamics in response to sensory, motor, or cognitive activation. Eight volunteers participated in the study where simultaneous EEG and 16 optode fNIR from anterior prefrontal cortex were recorded while participants engaged with the P300-BCI for spatial navigation. The results showed a significant response in fNIR signals during high, medium and low performance indicating a positive correlation between prefrontal oxygenation changes and BCI performance. This preliminary study provided evidence that the performance of P300-BCI can be monitored by fNIR which in turn can help improve the robustness of the BCI classification.

A P300-based EEG-BCI for spatial navigation control

In this study, a Brain Computer Interface (BCI) based on the P300 oddball paradigm has been developed for spatial navigation control in virtual environments. Functionality and efficacy of the system were analyzed with results from nine healthy volunteers. Each participant was asked to gaze at an individual target in a 3×3 P300 matrix containing different symbolic navigational icons while EEG signals were collected. Resulting ERPs were processed online and classification commands were executed to control spatial movements within the MazeSuite virtual environment and presented to the user online during an experiment. Subjects demonstrated on average, ~89% online accuracy for simple mazes and ~82% online accuracy in longer more complex mazes. Results suggest that this BCI setup enables guided free-form navigation in virtual 3D environments.

Detection of attention shift for asynchronous P300-based BCI

Brain-computer interface (BCI) provides patients suffering from severe neuromuscular disorders an alternative way of interacting with the outside world. The P300-based BCI is among the most popular paradigms in the field and most current versions operate in synchronous mode and assume participant engagement throughout operation. In this study, we demonstrate a new approach for assessment of user engagement through a hybrid classification of ERP and band power features of EEG signals that could allow building asynchronous BCIs. EEG signals from nine electrode locations were recorded from nine participants during controlled engagement conditions when subjects were either engaged with the P3speller task or not attending. Statistical analysis of band power showed that there were significant contrasts of attending only for the delta and beta bands as indicators of features for user attendance classification. A hybrid classifier using ERP scores and band power features yielded the best overall performance of 0.98 in terms of the area under the ROC curve (AUC). Results indicate that band powers can provide additional discriminant information to the ERP for user attention detection and this combined approach can be used to assess user engagement for each stimulus sequence during BCI use.

Brain in the Loop Learning Using Functional Near Infrared Spectroscopy

The role of practice is crucial in the skill acquisition process and for assessments of learning. In this study, we used a portable neuroimaging technique, functional near infrared (fNIR) spectroscopy for monitoring prefrontal cortex activation during learning of spatial navigation tasks throughout 11 days of training and testing. Two different tasks orders, blocked and random, were used to test the effect of the practice schedule on the acquisition and transfer of 3D computer mazes. Results indicated variable decreases in the hemodynamic response during the initial days of practice. Although there were no differences in mean oxygenation for the practice orders across acquisition the random practice order used less oxygenation than the blocked order for the more difficult tasks in the transfer phase Use of brain activation and behavioral measures provides can provide a more accurate depiction of the learning process. Since fNIR systems are safe, portable and record brain activation in ecologically valid settings, fNIR can contribute to future learning settings for assessment and personalization of the training regimen.

Differential Prefrontal Response during Natural and Synthetic Speech Perception: An fNIR Based Neuroergonomics Study

Synthetic speech has a growing role in human computer interaction and automated systems with the emergence of ubiquitous computing such as smart phones, car multimedia control and navigation systems. Cognitive processing costs associated with comprehension of synthetic speech relative to comprehension of natural speech have been demonstrated with behavioral (reaction time, accuracy, etc.) and self-reported (ratings, etc.) measures. In this neuroergonomics study, we have used optical brain imaging (fNIR: functional near infrared spectroscopy) to capture the brain activation of participants while they were listening to speech with varied quality, as well as natural speech. Results indicated a differential hemodynamic response with speech quality. As fNIR systems are safe, portable and record brain activation in real world settings, fNIR is a practical and minimally intrusive assessment tool for user experience researchers and can provide an objective metric for the design and development of next generation synthetic speech systems.

Functional brain activity monitoring during Unmanned Aerial Vehicle coordination

The efficiency and safety of many complex human-machine systems such as Unmanned Aerial Vehicle (UAV) are closely related to the cognitive workload and situational awareness of their operators. Subjective operator reports, physiological and behavioral measures are not reliably sensitive for monitoring cognitive overload. Drexels Optical Brain Imaging Team has already developed a portable safe and cost-effective optical brain imaging method called functional near-infrared spectroscopy (fNIR) for monitoring the prefrontal cortex in clinical and field settings. The current study examined the relationship of the hemodynamic response in prefrontal cortex to mental workload, level of expertise, and task performance.

Abnormal auditory-evoked gamma band oscillations in first-episode schizophrenia during both eye open and eye close states

&NA; Abnormal auditory steady state response (ASSR) is a typical finding among schizophrenia patients, which is thought to directly reflect deficient gamma band oscillations in the brain. However, whether these ASSR alterations are state dependent, e.g. during eye‐open or eye‐closed conditions, has not yet been carefully elucidated in schizophrenia. Our study aimed to explore whether the abnormality of ASSR in patients with first‐episode schizophrenia (FEP) is altered under eye‐open (EO) and eye‐closed (EC) states. ASSR was elicited using 40 Hz click trains under EO and EC states. Twenty‐eight healthy control subjects (HC) and thirty‐three FEP individuals, 17 of whom were medication‐naïve, were recruited. The event‐related spectrum perturbation (ERSP) and intertrial coherence (ITC) in response to 40 Hz click sounds were quantified. Compared to HC group, FEP group showed a lower ITC and ERSP during EO state, as well as a decreased ITC during EC state. Our results suggest that abnormalities in gamma band oscillations among first‐episode schizophrenia patients are present under both eye open and eye close states. Although differences in gamma band oscillations between EO and EC states within the FEP group were not observed, exploratory results suggest that state‐sensitivity may be contingent on medication use.

Increased resting-state global functional connectivity density of default mode network in schizophrenia subjects treated with electroconvulsive therapy

Modified electroconvulsive therapy (MECT) has been widely applied to help treat schizophrenia patients who are treatment-resistant to pharmaceutical therapy. Although the technique is increasingly prevalent, the underlying neural mechanisms have not been well clarified. We conducted a longitudinal study to investigate the alteration of global functional connectivity density (gFCD) in schizophrenia patients undergoing MECT using resting state fMRI (functional magnetic resonance imaging). Two groups of schizophrenia inpatients were recruited. One group received a four-week MECT together with antipsychotic drugs (ECT+Drug, n=21); the other group only received antipsychotic drugs (Drug, n=21). Both groups were compared to a sample of healthy controls (HC, n=23). fMRI scans were obtained from the schizophrenia patients twice at baseline (t1) and after 4-week treatment (t2), and from healthy controls at baseline. gFCD was computed using resting state fMRI. Repeated ANCOVA showed a significant interaction effect of group×time in the schizophrenia patients in left precuneus (Pcu), ventral medial prefrontal cortex (vMPFC), and dorsal medial prefrontal cortex (dMPFC) (GRF-corrected P<0.05), which are mainly located within the default mode network (DMN). Post-hoc analysis revealed that compared with baseline (t1), an increased gFCD was found in the ECT+Drug group in the dMPFC (t=3.87, p=0.00095), vMPFC (t=3.95, p=0.00079) and left Pcu (t=3.33, p=0.0034), but no significant effect was identified in the Drug group. The results suggested that increased global functional connectivity density within the DMN might be one important neural mechanism of MECT in schizophrenia.

Comparison of Machine Learning Approaches for Motor Imagery Based Optical Brain Computer Interface

A Brain-computer Interface (BCI) is a system that interprets specific patterns in human brain activity, such as the intention to perform motor functions, in order to generate a signal which can be used for communication or control. Functional near infrared spectroscopy (fNIRS) is an emerging optical neuroimaging technique which is a relatively new modality for BCI systems. As such, the optimal paradigms and classification techniques for the interpretation of fNIRS-BCI systems is an area of active investigation. Presently, most fNIRS BCIs have adopted Linear Discriminant Analysis (LDA) algorithm as the primary classification approach, however other alternative methods may offer increased performance. In order to compare different algorithms, a dataset from a four-class motor imagery-based fNIRS-BCI study was re-analyzed, and we systematically compared the performance of different machine learning algorithms: Naive Bayes (NB), LDA, Logistic Regression (LR), Support Vector Machines (SVM) and Multi-layer Perception (MLP). Our findings suggest that the LR classifier slightly outperformed other classifiers, unlike most fNIRS-BCI studies which reported LDA or SVM as the best classifier. The results presented here suggest that an LR classifier could be a potential replacement for LDA classifiers in motor imagery tasks.