Neurologic disorders frequently affect everyday functioning. However, clinicians and researchers have had limited success in predicting who will, or will not, succeed at real-world tasks, since most assessments rely on testing from in a controlled environment (clinic, laboratory) to predict performance in the less controlled and more chaotic real world. Researchers have therefore sought to develop EEG-based cognitive state algorithms to assess real-time neurophysiology during everyday activities. However, algorithms developed thus far have failed to generalize to complex real-world performance (e.g., on-road driving) and are rarely, if ever, validated in neurologic populations. The current proposal seeks to advance the B-Alert(R) X10 system (a portable, wireless EEG/ECG headset with cognitive state algorithms of engagement and workload) to address these limitations.
The aims of this project are: 1) to enhance the software and hardware capabilities of the ABM EEG system for use during on-road evaluations, 2) to validate EEG engagement and workload algorithms in individuals with HIV-associated neurocognitive disorders (HAND) in a laboratory-based environment, and 3) to relate EEG algorithms acquired during standardized on-road evaluations to driving performance, within control and HAND groups. HAND is an important population because 1) HIV is now a chronic condition, initially affecting individuals in young/middle age, 2) HAND remains prevalent and affects everyday functioning, including automobile driving, and 3) as with other conditions, clinic-based NP assessments only modestly predict success or failure at real world tasks. The project will occur in 3 phases. First, ABM will use a small set of on-road data already acquired to optimize both the cognitive state and artifact rejection algorithms. Next, the algorithms will be applied to in-laboratory cognitive and semi-naturalistic driving simulator assessments to determine what further adaptations might be required. Finally, participants will complete both an in-laboratory and standardized on-road assessment to determine the validity of the algorithms. Participants will be recruited from the HIV Neurobehavioral Research Center at UCSD, which will provide standardized neuropsychological and neuromedical assessments prior to enrollment. If successful, the proposed tool would 1) provide researchers with a new method for assessing components of real-world functioning, 2) validate the first on- road driving cognitive state algorithms, and 3) develop predictors of on-road driving impairments, reducing the need for potentially dangerous real-life driving assessments, The envisioned final product would allow for integration of in-laboratory and in-the-wild assessments for a variety of real world applications, and be available for use by researchers, clinicians, and public safety officials, and relevant to a broad range of conditions, including aging and various neurologic (e.g., stroke, TBI recovery) and psychiatric (e.g., substance use) disorders.
This project seeks to address the primary limitations of currently available cognitive state algorithms: failure to generalize to real-world performance outside of the laboratory, and failure to validate in neurologically impaired populations, particularly in HIV associated neurocognitive disorders (HAND). We propose to advance the portable, wireless EEG system, the B-Alert(R) X10 to accommodate these limitations, by adaptation of both the artifact decontamination and cognitive state (engagement and workload) algorithms.
