Neurological disorders affect hundreds of millions globally, yet translating artificial intelligence (AI) advances into clinical practice remains challenging due to fragmented, privacy-sensitive datasets. Federated learning (FL) has emerged as a promising paradigm, enabling collaborative model training across institutions without sharing raw patient data. This review synthesizes FL applications in neurology from 2020 to 2025, spanning neuroimaging, electrophysiology, and electronic health records. We analyze real-world deployments, highlight algorithmic trends, and discuss technical, regulatory, and organizational barriers to clinical translation. While FL demonstrates feasibility in tasks such as brain tumor segmentation, multiple sclerosis lesion detection, and electronic health record-based predictive modeling, verified clinical implementations remain scarce. We outline strategies to enhance adoption, including privacy-preserving techniques, standardized infrastructures, domain-adaptive algorithms, and cross-disciplinary collaboration. By bridging technical innovation with regulatory compliance and operational scalability, FL holds significant potential to advance precision neurology while safeguarding patient privacy.
Functional neuroimaging has revolutionized our understanding of brain physiology and pathophysiology, providing dynamic insights into neural activity that complement structural imaging. This review examines the principal functional neuroimaging modalities used in routine clinical practice, positron emission tomography (PET), single-photon emission computed tomography, magnetic resonance imaging, and their clinical applications in neurology. We present an evidence-based approach to modality selection and interpretation, emphasizing the integration of multimodal imaging data to enhance diagnostic accuracy. Recent advances in molecular imaging have transformed the evaluation of neurodegenerative diseases, moving the field from observing the downstream consequences of disease to visualizing the primary molecular pathologies in vivo. The development of specific PET ligands for amyloid-β plaques, tau neurofibrillary tangles, neurotransmitter system components, and markers of synaptic density and neuroinflammation represents a paradigm shift toward a biological definition of these disorders. This review provides practical algorithms for clinical decision-making, critically evaluates the strengths and limitations of each technique, and highlights emerging applications that promise to further advance neurological diagnosis and management toward an era of precision medicine.
Child neurology training has undergone minimal change over the decades, despite a rapid growth in subspecialty knowledge, patient volumes, and complexity. The current 5-year structure, which was established due to necessary historical compromises between pediatrics and neurology, is increasingly misaligned with modern clinical practice and educational priorities. Most child neurologists no longer pursue dual pediatrics certification, and few provide neurologic care to adult patients. Meanwhile, the field has expanded significantly in complexity and volume, making it a large enough specialty to sustain an independent curriculum. We propose a streamlined 4-year categorical residency model that integrates relevant components of pediatrics and adult neurology while centering training around child neurology from the start. This model, which aligns better with structures seen in comparable specialties, prioritizes flexibility and increases the opportunities for longitudinal mentorship and professional development. Thoughtful planning and collaboration will be essential to surmount challenges during the transition, including changes in board certification and alterations to institutional funding. Modernizing child neurology training is essential to better prepare future specialists, support recruitment and resident development, and meet the evolving needs of children with neurologic disorders.
Over the last decade, there has been a rapid expansion in medical devices utilizing artificial intelligence (AI) and machine learning (ML), with a growing role in neurologic care. These devices are beginning to augment clinical workflows and reshape how neurologists engage with technology to deliver patient care. In this review, we first introduce core ML techniques that are used within devices. We then describe the AI-enabled medical devices that have received U.S. Food and Drug Administration authorization as of December 31, 2024, including an analysis of the 147 devices across neuroradiology and broader neurology indications. We also highlight key trends in how these devices are being integrated into clinical practice. We conclude by examining emerging models of human-machine interaction and their implications for future neurologic care.
Common carotid artery (CCA) disease represents a challenging subset of cerebrovascular pathology with unique anatomical and hemodynamic considerations distinct from the more extensively studied internal carotid artery stenosis. Stenosis at the CCA origin from the aortic arch occurs with an approximate incidence of 0.5 to 6.4% of angiographic imaging cases, with approximately 17% of symptomatic patients developing tandem disease involving the carotid bifurcation. The pathophysiology of stroke risk involves hemodynamic compromise, embolic phenomena, and the cumulative effects of multilevel disease, although natural history data remain limited. Management decisions are guided primarily by observational data and expert consensus due to the absence of randomized controlled trials. Current evidence-based guidelines recommend optimal medical management for asymptomatic patients and reserve intervention for symptomatic individuals. When intervention is pursued, an "endovascular-first" approach is favored based on systematic review data showing lower 30-day mortality and stroke rates compared with open surgery, despite higher restenosis rates (9 vs. 1.3%). However, endovascular treatment of tandem lesions increases stroke and death risk in asymptomatic patients. Hybrid procedures combining open and endovascular techniques demonstrate intermediate outcomes but elevated risk in symptomatic patients. Treatment selection requires individualized assessment incorporating symptom status, lesion characteristics, technical feasibility, patient anatomy, and surgeon experience. Substantial knowledge gaps persist regarding natural history, long-term comparative effectiveness, optimal management of tandem disease, and the evolving role of contemporary medical therapy, underscoring the need for continued investigation in this complex area of vascular medicine.
As the U.S. population ages, Alzheimer's disease and related dementias (ADRD) cases are increasing, resulting in long wait times for specialist care. We review state-of-the-art artificial intelligence (AI) applications in ADRD care, from streamlining clinical diagnosis to pioneering novel digital biomarkers. Near-term AI applications include neuroimaging interpretation, conversational agents for patient interviews, and digital cognitive assessments. Large language models show promise as collaborative partners, helping clinicians interpret complex data while supporting patients and caregivers. Emerging digital biomarkers-speech analysis, passive monitoring through wearable devices, electronic health record analysis, and multiomics-offer potential for continuous monitoring to detect cognitive decline years before traditional assessments. Despite the acceleration of AI innovation, most of these systems are inaccessible in clinical practice. Implementation bottlenecks include limited external validation, technical challenges, model biases, infrastructure, and regulatory requirements. This review aims to help neurologists navigate this rapidly evolving AI landscape and prepare for implementation in ADRD care.
Artificial intelligence (AI) is rapidly transforming clinical practice, necessitating that neurology educators prepare trainees for this shift. This review outlines how AI tools are currently being trialed for integration into clinical care and demonstrates potential applications within neurology education. We highlight practical models for curriculum development from other specialties and underscore the need for structured, competency-driven approaches. Ultimately, integrating AI education into neurology training is essential to equip future clinicians for a healthcare system increasingly shaped by AI.
Carotid atherosclerosis is a major cause of ischemic stroke, historically managed according to luminal stenosis severity. However, stenosis alone fails to capture plaque biology, as features such as intraplaque hemorrhage (IPH), lipid-rich necrotic core, fibrous cap rupture, ulceration, and perivascular inflammation strongly influence vulnerability and clinical outcomes. Advances in imaging have shifted the focus from lumen quantification to multimodal plaque phenotyping. Ultrasound, computed tomography angiography (CTA), magnetic resonance imaging (MRI), digital subtraction angiography, photon-counting CT (PCCT), and positron emission tomography (PET) provide complementary morphological and biological information, while contrast-enhanced ultrasound, radiomics, and artificial intelligence (AI) enhance risk stratification beyond conventional parameters. MRI remains the reference for detecting IPH, while CTA and PCCT improve lumen and tissue characterization. PET tracers further interrogate inflammatory and calcification pathways. Emerging frameworks such as Plaque-Reporting and Data System aim to standardize reporting and integrate multimodal biomarkers. The future of carotid imaging lies in precision medicine, combining advanced imaging, AI, and molecular biology to optimize individualized stroke prevention strategies.
The preclinical neuroscience course is widely regarded as one of the most conceptually challenging components of medical education and is often associated with the emergence of "neurophobia." Rather than continuing to focus solely on reducing fear, it is time to cultivate "neuro-curiosity" a mindset of inquiry, relevance, and connection. Grounded in self-determination theory, which emphasizes autonomy, competence, and relatedness as drivers of intrinsic motivation, this review outlines four core pillars for curriculum reform: clinical relevance, neuroanatomy through imaging, case-based learning, and digital engagement. Together, these strategies promote deeper learning, emotional engagement, and diagnostic reasoning. With neurologic disease burden rising and a projected shortage of neurologists, early, engaging exposure to neuroscience is critical to building a robust future workforce. The preclinical neuroscience course presents a unique opportunity not only to teach foundational knowledge, but also to inspire sustained interest in neurology from the very start of medical training.
As a field, neurology can seem complicated, overwhelming, and at times ambiguous and uncertain. However, novel technological developments-including artificial intelligence-can be used to decrease "neurophobia," foster enthusiasm about our specialty, and increase curiosity and motivation while decreasing the educator's time in preparation. This review discusses the technology-enhanced teaching landscape and how neurology integrates into known conceptual frameworks and learning theories in medical education. We provide detailed guidance for technology-focused curriculum design in all realms of neurologic teaching: at the bedside, in small groups, and in larger presentations. Finally, we identify ways technological scholarship can be leveraged toward academic promotion for neurology education.
The role of the neurologic educator has evolved from an informal pursuit into an intentional academic discipline. This development intersects with the persistent challenge of neurophobia, characterized by fear and avoidance of neurology among trainees. Despite improved understanding of its drivers, neurophobia continues to negatively influence engagement and recruitment into the field. Importantly, many contributing factors reflect modifiable features of educational design rather than inherent characteristics of neurology, creating opportunities for educational innovation. Here, we situate the modern neurologic educator within the evolution of medical education and describe how the complementary roles of teacher, administrator, and scholar can advance neurologic education and address neurophobia. We also highlight opportunities in neurologic education, including engagement with nonphysician clinicians, use of nontraditional platforms for dissemination, and a shift toward cultivating neurophilia. Central to this work is the intentional creation of career pathways that inspire, support, and sustain the next generation of neurologic educators.
We describe efforts to systematically improve care delivery and outcomes in patients with functional neurological disorder (FND) via the creation of a program that allows the delivery of comprehensive clinical care, backed by administrative support. Collaborative efforts are necessary to develop an efficient, effective, long-term, and sustainable program. The creation of multi-site clinic service agreements, electronic medical records templates, multidisciplinary meetings, and trainee and staff education aids in guaranteeing the reliability and impact on patient care, outcomes, and knowledge in the FND field. We review barriers encountered in the process of creating a program and describe solutions to enable a system that builds bridges to integrate whole-person care for patients with FND. The success of a multidisciplinary program relies on a systems approach, administrative coordination, team communication, defined roles, and initiative and oversight by a site champion, for a focus on each patient as an individual and on continuity of care. The FND program also established an educational path to train clinicians and encourage interdisciplinary clinical research in the field and programmatic evaluation.
The widespread adoption of virtual residency interviews in response to the COVID-19 pandemic led to an explosion in literature comparing the pros and cons of virtual and in-person interviews, but also led to an explosion in already-high residency application and interview volumes. While virtual interviews were substantially cheaper for all involved, there is fear that applicants and programs cannot judge one another as well as during in-person interviews. Likewise, increases in application volumes have made holistic application review more challenging for program directors, but the recent rise in "preference signaling" seems to be an optimal solution to that issue. 2020 also saw increased awareness of systemic inequities in the United States, and medical education and residency recruitment was not immune from scrutiny. Finally, the rise of artificial intelligence could again fundamentally change the resident selection process. It is imperative that the GME community continues to adapt to a changing world.
Acute stroke alerts are frequently triggered by conditions unrelated to cerebrovascular disease, resulting in false positives that burden clinical teams and contribute to diagnostic ambiguity. At a large academic center, we developed ScanNER v2, a machine learning (ML) model based on large-language models (LLMs) and structured clinical data to predict the presence of acute cerebrovascular disease (ACD) in approximately 16,000 stroke alerts occurring over 10 years with an area under the receiver-operating curve and F1 score of 0.72 and overall positive predictive value of 0.68. In this perspective, we outline a practical framework for operationalizing this model within hospital-based stroke systems. We first describe our health-system experience developing and validating an AI-enabled pipeline, named "ScanNER 2," then take the point of view of two implementation angles (high sensitivity and high specificity), outlining the operational and clinical tradeoffs for each approach. We also highlight challenges related to implementation, clinical governance, workflow integration, and equity, emphasizing guardrails required for responsible deployment. As stroke centers increasingly adopt AI-assisted tools, this type of thought experiment is essential to ensure that such ML-based innovations effectively enhance the core mission of delivering timely, high-quality acute stroke care.
Links between the eye and the central nervous system (CNS) have been recognized since the origins of the ophthalmoscope. Owing to the elegant topography of the afferent visual pathway and its close embryonic, anatomical, and physiological connections to the brain, it is possible to capture structural effects of CNS injury in the retina. The availability of large-scale, high-quality retinal imaging datasets and ongoing advances in artificial intelligence (AI) have paved the way for Oculomics, a field in which ocular measures act as biomarkers for systemic diseases. Similarly, ocular images have been used in AI models to provide critical insights about neurologic disorders in the fledgling discipline of what might be considered Neuro-Oculomics. In this review, we will describe key ocular imaging techniques and highlight emerging roles for AI in the diagnosis and management of important neurological conditions.
There is growing evidence supporting specific approaches to treat functional neurological disorder (FND), which commonly employ a multidisciplinary strategy guided by the biopsychosocial model and holistic perspective. Integrative Medicine and Health is an emerging specialty that brings together conventional and complementary modalities to deliver coordinated and comprehensive care. Although integrative approaches can be incorporated into existing care models, their application in the treatment of FND has not been systematically investigated, and programs employing such approaches to FND remain exceedingly rare. This study outlines the potential role of integrative care in FND. It will characterize the integrative approach and highlight its potential benefits for FND, review current evidence for relevant therapies, propose a potential clinic workflow, and illustrate its application through clinical case vignettes.
The utility of neuroimaging in the diagnosis and management of movement disorders has been steadily increasing as both imaging and image analysis technologies have advanced in the last decade. Neuroimaging is also playing a critical role in the search for novel therapies to prevent, slow down, and treat various movement disorders. This article reviews both standard and innovative imaging tools available for both clinicians and researchers. We focus predominantly on the clinician's perspective, discussing imaging tools that are becoming rapidly available and how these may be integrated into the clinic to provide cutting-edge and patient-centered care. We discuss novel and emerging techniques and their potential implications for the field, as well as highlight areas still in need of imaging solutions.
Persistent symptoms after concussion (PSaC) and functional neurological disorder (FND) are frequently encountered in clinical practice and are often challenging to manage due to heterogeneous and polysymptomatic presentations, as well as fragmented care pathways. This review outlines key points of intersection between PSaC and FND across pathophysiology, illness beliefs, rehabilitation models, and emerging treatments. We describe when FND should be considered in the differential diagnosis of patients with PSaC, and provide guidance on history-taking, examination, diagnostic communication, and rehabilitation planning that can be applied to both conditions. We also examine the influence of expectations, clinical messaging, and interactions with the healthcare system on recovery. Integrating principles from FND into concussion care may help clinicians more accurately formulate cases and support individualized rehabilitation pathways.
Essential tremor (ET) is the most common cause of tremor worldwide and can become profoundly disabling in many patients, with pharmacological treatments often providing insufficient relief. Surgical interventions have emerged as effective strategies for long-term tremor control. This review summarizes the current evidence on surgical therapies, including deep brain stimulation (DBS), radiofrequency (RF) thalamotomy, magnetic resonance-guided focused ultrasound (MRgFUS), and Gamma Knife radiosurgery (GKSR) for ET and other tremor-inducing syndromes. These techniques demonstrate comparable efficacy. DBS offers the advantage of adjustable parameters, allowing optimization of the therapeutic window while minimizing adverse effects. MRgFUS is particularly attractive due to its minimally invasive nature, whereas RF thalamotomy and GKSR remain viable alternatives for patients who are ineligible for DBS or MRgFUS. Bilateral interventions are increasingly feasible, and treatment selection should be individualized, considering clinical characteristics and patient preference. Ongoing advances in magnetic resonance imaging (MRI) technology and neurostimulation are poised to further refine surgical management and improve outcomes for patients with tremor.
Functional tremor (FT) is the most prevalent subtype of functional movement disorders, characterized by variability, distractibility, and entrainability on clinical examination. Diagnosis relies on positive clinical and electrophysiological signs, shifting emphasis from a diagnosis of exclusion to "rule-in" criteria. Surface electromyography and tremor analysis are essential tools in establishing the diagnosis. Pathophysiology involves abnormal motor co-activation, disrupted volitional awareness, and impaired predictive processing, resulting in tremor perceived as involuntary despite intact motor pathways. Management requires a multidisciplinary approach, including physiotherapy, occupational therapy, cognitive-behavioral interventions, biofeedback, and transcranial magnetic stimulation. FT often results in persistent disability, with limited treatment response. Management is hindered by diagnostic challenges, especially in functional overlay, limited training, cultural misconceptions, and underutilization of neurophysiological and rehabilitation interventions. Improving clinician training, expanding access to neurophysiology, and multidisciplinary care, along with high-quality prospective specific research and standardized care pathways, is essential to optimize outcomes.