Evaluating the Test-Retest Reliability of Multimodal Brain Imaging

Abstract

Test-retest reliability is essential in research to ensure that neuroimaging devices produce reliable outcomes, especially prior to their adoption in clinical settings. Simultaneous fNIRS-EEG is a growing multimodal neuroimaging modality used to study brain function, combining fNIRS to measure the hemodynamic response with EEG to measure the electrical neuronal activity. This multimodal approach allows for the assessment of neurovascular coupling, a mechanism in the brain that connects neural activity with cerebral blood flow to maintain proper brain function. While previous studies have encountered challenges, my thesis work has demonstrated that this multimodal approach can reliably assess neurovascular coupling across multiple brain regions and cognitive task stimuli. Establishing the test-retest reliability of simultaneous fNIRS-EEG provides the opportunity to assess clinical populations. My thesis has utilized the fNIRS-EEG protocol to study patients with mesial temporal lobe epilepsy and evaluated their neurovascular coupling abnormalities. In line with literature, this simultaneous device identified abnormally uncoupled neurovascular responses in epileptic patients in comparison with the healthy subjects, suggesting that fNIRS-EEG may serve as a potential pre-assessment tool in the clinical workflow prior to language and memory lateralization exams and surgical resection. Building upon the previously established system, my thesis also evaluates the reliability of a state-of-the-art multimodal device, concurrent fNIRS, EEG and transcranial magnetic stimulation (fNET), which promises a new treatment for medication-resistant depression. My evaluation using a previously established standardized protocol has found that the motor responses from the newly integrated fNET system still demonstrated consistent test-retest reliability as seen before, even with whole-head coverage. However, the responses to single pulse transcranial magnetic stimulation are variable across individuals, while intra-individual-level responses are more consistent. Further research is warranted to establish group-level reliability. Overall, this thesis demonstrates the use of multimodal approaches in brain imaging to reliably assess healthy and clinical patient populations. Furthermore, the integration of fNIRS-EEG with TMS provides the opportunity to increase spatial targeting of TMS and advance treatments for depression and other neurological disorders.