Faculty Bibliography

BACKGROUND:Neuro-ophthalmology frequently requires a complex and multi-faceted clinical assessment supported by sophisticated imaging techniques in order to assess disease status. The current approach to diagnosis requires substantial expertise and time. The emergence of AI has brought forth innovative solutions to streamline and enhance this diagnostic process, which is especially valuable given the shortage of neuro-ophthalmologists. Machine learning algorithms, in particular, have demonstrated significant potential in interpreting imaging data, identifying subtle patterns, and aiding clinicians in making more accurate and timely diagnosis while also supplementing nonspecialist evaluations of neuro-ophthalmic disease. EVIDENCE ACQUISITION/METHODS:Electronic searches of published literature were conducted using PubMed and Google Scholar. A comprehensive search of the following terms was conducted within the Journal of Neuro-Ophthalmology: AI, artificial intelligence, machine learning, deep learning, natural language processing, computer vision, large language models, and generative AI. RESULTS:This review aims to provide a comprehensive overview of the evolving landscape of AI applications in neuro-ophthalmology. It will delve into the diverse applications of AI, optical coherence tomography (OCT), and fundus photography to the development of predictive models for disease progression. Additionally, the review will explore the integration of generative AI into neuro-ophthalmic education and clinical practice. CONCLUSIONS:We review the current state of AI in neuro-ophthalmology and its potentially transformative impact. The inclusion of AI in neuro-ophthalmic practice and research not only holds promise for improving diagnostic accuracy but also opens avenues for novel therapeutic interventions. We emphasize its potential to improve access to scarce subspecialty resources while examining the current challenges associated with the integration of AI into clinical practice and research.

Cenobamate has demonstrated efficacy in patients with treatment-resistant epilepsy, including patients who continued to have seizures after epilepsy surgery. This article provides recommendations for cenobamate use in patients referred for epilepsy surgery evaluation. A panel of six senior epileptologists from the United States and Europe with experience in presurgical evaluation of patients with epilepsy and in the use of antiseizure medications (ASMs) was convened to provide consensus recommendations for the use of cenobamate in patients referred for epilepsy surgery evaluation. Many patients referred for surgical evaluation may benefit from ASM optimization; both ASM and surgical treatment should be individualized. Based on previous clinical studies and the authors' clinical experience with cenobamate, a substantial proportion of patients with treatment-resistant epilepsy can become seizure-free with cenobamate. We recommend a cenobamate trial and ASM optimization in parallel with presurgical evaluations. Cenobamate can be started before phase two monitoring, especially in patients who are found to be suboptimal surgery candidates. As neurostimulation therapies are generally palliative, we recommend trying cenobamate before vagus nerve stimulation (VNS), deep brain stimulation, or responsive neurostimulation (RNS). In surgically remediable cases (mesial temporal sclerosis, benign discrete lesion in non-eloquent cortex, cavernous angioma, etc.), cenobamate use should not delay imminent surgery; however, a patient may decide to defer or even cancel surgery should they achieve sustained seizure freedom with cenobamate. This decision should be made on an individual, case-by-case basis based on seizure etiology, patient preferences, potential surgical risks (mortality and morbidity), and likely surgical outcome. The addition of cenobamate after unsuccessful surgery or palliative neuromodulation may also be associated with better outcomes.

Cases of isolated spinal cord ischemia resulting in symptoms in neonates are rare, and there are even fewer reported cases in atraumatic births. We present a case of a presumed isolated cervical cord ischemic injury, discuss differentials to consider when evaluating a neonatal spinal cord injury, and highlight the difficulties of diagnosing a spinal cord infarction.

Alzheimer's disease (AD) is a devastating, age-associated neurodegenerative disorder and the most common cause of dementia. The clinical continuum of AD spans from preclinical disease to subjective cognitive decline, mild cognitive impairment, and dementia stages (mild, moderate, and severe). Neuropathologically, AD is defined by the accumulation of amyloid β (Aβ) into extracellular plaques in the brain parenchyma and in the cerebral vasculature, and by abnormally phosphorylated tau that accumulates intraneuronally forming neurofibrillary tangles (NFTs). Development of treatment approaches that prevent or even reduce the cognitive decline because of AD has been slow compared to other major causes of death. Recently, the United States Food and Drug Administration gave full approval to 2 different Aβ-targeting monoclonal antibodies. However, this breakthrough disease modifying approach only applies to a limited subset of patients in the AD continuum and there are stringent eligibility criteria. Furthermore, these approaches do not prevent progression of disease, because other AD-related pathologies, such as NFTs, are not directly targeted. A non-mutually exclusive alternative is to address lifestyle interventions that can help reduce the risk of AD and AD-related dementias (ADRD). It is estimated that addressing such modifiable risk factors could potentially delay up to 40% of AD/ADRD cases. In this review, we discuss some of the many modifiable risk factors that may be associated with prevention of AD/ADRD and/or increasing brain resilience, as well as other factors that may interact with these modifiable risk factors to influence AD/ADRD progression. ANN NEUROL 2024.

BACKGROUND AND OBJECTIVES/UNASSIGNED:Exposure to repetitive head impacts (RHI) is linked to the development of chronic traumatic encephalopathy (CTE), which can only be diagnosed at post-mortem. The presence of a cavum septum pellucidum (CSP) is a common finding in post-mortem studies of confirmed CTE and in neuroimaging studies of individuals exposed to RHI. This study examines CSP in living former American football players, investigating its association with RHI exposure, traumatic encephalopathy syndrome (TES) diagnosis, and provisional levels of certainty for CTE pathology. METHODS/UNASSIGNED:Data from the DIAGNOSE CTE Research Project were used to compare the presence and ratio of CSP in former American football players (n = 175), consisting of former college (n = 58) and former professional players (n = 117), and asymptomatic unexposed controls without RHI exposure (n = 55). We further evaluated potential associations between CSP measures and cumulative head impact index (CHII) measures (frequency, linear acceleration, and rotational force), a TES diagnosis (yes/no), and a provisional level of certainty for CTE pathology (suggestive, possible, and probable). RESULTS/UNASSIGNED:Former American football players exhibited a higher CSP presence and ratio than unexposed asymptomatic controls. Among player subgroups, professional players showed a greater CSP ratio than former college players and unexposed asymptomatic controls. Among all football players, CHII rotational forces correlated with an increased CSP ratio. No significant associations were found between CSP measures and diagnosis of TES or provisional levels of certainty for CTE pathology. DISCUSSION/UNASSIGNED:This study confirms previous findings, highlighting a greater prevalence of CSP and a greater CSP ratio in former American football players compared with unexposed asymptomatic controls. In addition, former professional players showed a greater CSP ratio than college players. Moreover, the relationship between estimates of CHII rotational forces and CSP measures suggests that cumulative frequency and strength of rotational forces experienced in football are associated with CSP. However, CSP does not directly correlate with TES diagnosis or provisional levels of certainty for CTE, indicating that it may be a consequence of RHI associated with rotational forces. Further research, especially longitudinal studies, is needed for confirmation and to explore changes over time.

As neuropalliative care is better recognized and more widely utilized, there is as great a need for clinicians trained in the field as there is for disease-specific symptom management, advance care planning, and end-of-life care. In this manuscript, we describe potential career trajectories in neuropalliative care. For clinicians, this includes educational and training opportunities within primary neuropalliative care (integrating palliative care principles into usual neurology practice), specialty neuropalliative care (completing a hospice and palliative medicine fellowship), and hospice. We also describe considerations for establishing new clinical neuropalliative practices and highlight neuropalliative education and research as key areas for advancing the field.

Sensory inputs enter a constantly active brain, whose state is always changing from one moment to the next. Currently, little is known about how ongoing, spontaneous brain activity participates in online task processing. We employed 7 Tesla fMRI and a threshold-level visual perception task to probe the effects of prestimulus ongoing brain activity on perceptual decision-making and conscious recognition. Prestimulus activity originating from distributed brain regions, including visual cortices and regions of the default-mode and cingulo-opercular networks, exerted a diverse set of effects on the sensitivity and criterion of conscious recognition, and categorization performance. We further elucidate the mechanisms underlying these behavioral effects, revealing how prestimulus activity modulates multiple aspects of stimulus processing in highly specific and network-dependent manners. These findings reveal heretofore unknown network mechanisms underlying ongoing brain activity's influence on conscious perception, and may hold implications for understanding the precise roles of spontaneous activity in other brain functions.

OBJECTIVE/UNASSIGNED:This study investigates speech decoding from neural signals captured by intracranial electrodes. Most prior works can only work with electrodes on a 2D grid (i.e., Electrocorticographic or ECoG array) and data from a single patient. We aim to design a deep-learning model architecture that can accommodate both surface (ECoG) and depth (stereotactic EEG or sEEG) electrodes. The architecture should allow training on data from multiple participants with large variability in electrode placements and the trained model should perform well on participants unseen during training. APPROACH/UNASSIGNED:We propose a novel transformer-based model architecture named SwinTW that can work with arbitrarily positioned electrodes, by leveraging their 3D locations on the cortex rather than their positions on a 2D grid. We train both subject-specific models using data from a single participant as well as multi-patient models exploiting data from multiple participants. MAIN RESULTS/UNASSIGNED:The subject-specific models using only low-density 8x8 ECoG data achieved high decoding Pearson Correlation Coefficient with ground truth spectrogram (PCC=0.817), over N=43 participants, outperforming our prior convolutional ResNet model and the 3D Swin transformer model. Incorporating additional strip, depth, and grid electrodes available in each participant (N=39) led to further improvement (PCC=0.838). For participants with only sEEG electrodes (N=9), subject-specific models still enjoy comparable performance with an average PCC=0.798. The multi-subject models achieved high performance on unseen participants, with an average PCC=0.765 in leave-one-out cross-validation. SIGNIFICANCE/UNASSIGNED:The proposed SwinTW decoder enables future speech neuroprostheses to utilize any electrode placement that is clinically optimal or feasible for a particular participant, including using only depth electrodes, which are more routinely implanted in chronic neurosurgical procedures. Importantly, the generalizability of the multi-patient models suggests the exciting possibility of developing speech neuroprostheses for people with speech disability without relying on their own neural data for training, which is not always feasible.

Recent research has used large language models (LLMs) to study the neural basis of naturalistic language processing in the human brain. LLMs have rapidly grown in complexity, leading to improved language processing capabilities. However, neuroscience researchers haven't kept up with the quick progress in LLM development. Here, we utilized several families of transformer-based LLMs to investigate the relationship between model size and their ability to capture linguistic information in the human brain. Crucially, a subset of LLMs were trained on a fixed training set, enabling us to dissociate model size from architecture and training set size. We used electrocorticography (ECoG) to measure neural activity in epilepsy patients while they listened to a 30-minute naturalistic audio story. We fit electrode-wise encoding models using contextual embeddings extracted from each hidden layer of the LLMs to predict word-level neural signals. In line with prior work, we found that larger LLMs better capture the structure of natural language and better predict neural activity. We also found a log-linear relationship where the encoding performance peaks in relatively earlier layers as model size increases. We also observed variations in the best-performing layer across different brain regions, corresponding to an organized language processing hierarchy.

INTRODUCTION/UNASSIGNED:-variant carrier status will help target genetic testing in clinical settings where cost and access limit its availability. METHODS/UNASSIGNED:variant carrier status. The model was cross-validated across the two cohorts. RESULTS/UNASSIGNED:variants in the PPMI cohort and study cohort (AUC 0.897 and 0.738, respectively). The PPMI cohort model successfully generalized to the study cohort data using both MDS-UPDRS scores and binomial data (AUC 0.740 and 0.734, respectively) while the study cohort model did not. CONCLUSIONS/UNASSIGNED:variants.