About
Schedule
Speakers
Registration
October 30, 2025 (Thursday)
Session I - Cutting-edge techniques in advanced neuroimaging
Brain tissue microstructure imaging with diffusion MRI - the new frontier powered by Cima.X and AI
In brief
It was in 1965, exactly 60 years ago, when Stejskal and Tanner invented the magnetic resonance technique that we now call diffusion-weighted MRI or diffusion MRI for short. This technique has been of great interest because it allows us to assess tissue micro-structure – the characteristics of biological tissue at the micron scale. Diffusion MRI has proven particularly useful in the investigation of the living human brain both in health and disease. In this talk, I will give an overview of brain tissue microstructure imaging with diffusion MRI and share my thoughts on the exciting new possibilities Cima.X, coupled with AI, will bring.
TBD
Toward high resolution myelin imaging
In brief
Magnetic susceptibility imaging has progressed from SWI to QSM, enabling quantitative assessment of iron in the brain. However, QSM combines signals from multiple sources, primarily iron and myelin, which have opposite magnetic properties. In this talk, I will introduce our lab’s development of susceptibility source separation, a technique that disentangles these signals to independently map myelin and iron in vivo. I will discuss the biophysical modeling and validation behind the method, and highlight its potential applications in studying cortical myelination, multiple sclerosis, and other neurological conditions where myelin and iron play key roles.
Magnetic susceptibility imaging has evolved significantly since the introduction of susceptibility-weighted imaging (SWI), which was initially developed for clinical appli-cations such as detecting microhemorrhages. In the mid-2000s, advance-ments in high-resolution phase imaging led to the development of quantitative susceptibility mapping (QSM), a method that allows for the quantitative assessment of magnetic susceptibility in the brain. QSM has opened up new possibilities for neuroimaging, particularly in the quantification of iron in deep brain structures, a biomarker asso-ciated with various neurological conditions.
While QSM has provided valuable insights, it has primarily been limited to measuring a combined signal from different sources of magnetic susceptibility. In the brain, iron and myelin are the two dominant contributors, each with opposite magnetic properties—iron being paramagnetic and myelin being diamagnetic. This limitation has motivated the need for more refined techniques to disentangle these distinct sources of suscepti-bility.
In this presentation, I will introduce a technique developed in our lab called suscepti-bility source separation. This method provides a breakthrough by allowing the separate quantification of myelin and iron in the brain. Through sus-ceptibility source separation, we can now obtain high-resolution maps that independently profile the distribution of myelin and iron. This is particularly valuable for studying the brain, where myelin and iron play crucial roles in both healthy brain function and the progression of neurologi-cal diseases.
I will discuss the underlying physics that enables this separation, including the bio-physical modeling and algorithm employed. The validation of the method has been demonstrated through various approaches including histology, showing its accuracy in separating the signals of myelin and iron. The clinical potential of this technique is significant, with early studies indicating its applicability to diseases such as multiple sclerosis, where both iron deposition and myelin degra-dation are key pathological features. In addition, the method holds promise for more accurate mapping of cortical myelin-ation, which could improve our understanding of a variety of neurodegenerative diseases and brain aging.
Exploring Glymphatic Flow: Innovative MRI-Based Visualization Approaches
In brief
The glymphatic system plays a vital role in clearing metabolic waste from the brain, with dysfunction linked to aging and neurodegenerative disease. In this talk, I will present innovative MRI methods that allow us to non-invasively visualize and quantify glymphatic flow in vivo. I will highlight approaches such as diffusion tensor imaging, intrathecal contrast-enhanced MRI, and dynamic imaging protocols, which reveal the spatial and temporal dynamics of cerebrospinal and interstitial fluid transport. I will also discuss how glymphatic activity changes with physiological states like sleep and exercise, and how it is altered in disease models. Together, these methods provide new opportunities to study brain waste clearance and its clinical implications.
The glymphatic system has emerged as a critical facilitator of metabolic waste clearance in the central nervous system, with implications for neurodegenerative disease pathogenesis and brain homeostasis. Traditional methods for studying glymphatic function have relied on invasive procedures or exogenous tracers, limiting their utility for translational and clinical applications. Recent advances in magnetic resonance imaging (MRI) have enabled non-invasive assessment of glymphatic transport and provided unprecedented insights into the spatial and temporal dynamics of cerebrospinal fluid (CSF) and interstitial fluid (ISF) flow.
In this talk, I present innovative MRI methodologies to visualize and analyze glymphatic pathways in vivo, focusing on techniques such as diffusion tensor imaging (DTI), intrathecal contrast-enhanced MRI, and dynamic contrast-enhanced MRI. These approaches allow for quantitative and qualitative evaluation of glymphatic function, including tracer move-ment, perivascular transport, and regional differences in CSF-ISF exchange. Utilizing advanced image processing and modeling algorithms, we were able to delineate the peri-vascular spaces and measure fluid propagation through brain parenchyma with high spatial resolution. Our findings demonstrate that glymphatic flow patterns are markedly influenced by physiological states such as sleep, anesthesia, and exercise. Moreover, alterations in glymphatic transport were ob-served in models of aging and neurodegenerative conditions, highlighting the clinical relevance of MRI-based glymphatic visualization. The incorporation of dynamic imaging protocols enabled us to capture the real-time kinetics of solute clearance and revealed regional heterogeneity in glymphatic function across cortical and subcortical structures. Notably, these novel MRI techniques correlate with traditional histological and tracer-based measurements, validating their utility for preclinical and potential clinical investigation.
In conclusion, our study underscores the significance of innovative MRI approaches for elucidating glymphatic system physiology and pathophysiology. The ability to non-invasively visualize and quantify glymphatic pathways represents a major advancement in neuroimaging, with substantial implications for early diagnosis and monitoring of brain disorders associated with impaired waste clearance. These methods have the potential to accelerate translational research and guide therapeutic strategies targeting glymphatic dysfunction.
Imaging Brain Microstructure and Function with Ultra-High-Field MRI
In brief
Ultra-high-field 7 Tesla MRI has opened new possibilities for imaging brain structure and function with unprecedented detail. In this talk, I will highlight how 7T enables both functional and structural imaging, as well as quantitative neurochemical measurements with MR spectroscopy. I will also discuss the benefits of higher field strength, including stronger BOLD signals and improved contrast for visualizing fine brain microstructures such as cortical layers in vivo. Finally, I will share our work using 7T to achieve high-resolution human brain imaging and its implications for neuroscience research.
Magnetic resonance imaging (MRI) observation of the living brain depends on spatial resolution, signal-to-noise ratio (SNR), and tissue parameters such as relaxation time and contrast.
The advent of 7 Tesla (T) ultra-high-field MRI offers unprecedented capabilities for noninvasive imaging of human and animal brains. This technical capability spans a range of functional and structural domains and provides new oppor-tunities for quantitative neurochemical measurements using MR spectroscopic techniques.
Additionally, increasing the static magnetic field enhances phase dispersion and induces shifts in the signal, resulting in predicted benefits such as an enhanced Blood Oxygenation Level Dependent (BOLD) effect, which is used to detect brain activity, and improved contrast-to-noise ratio (CNR). Using optimal measurement techni-ques, improved CNR allows for the delineation of brain microstructures, including the laminar structures of the cerebral cortex in vivo.
In this presentation, I would like to introduce our research on 7T in-vivo high-resolution human brain imaging.
Large Brain Models for Neuroscience
In brief
This talk introduces the concept of Large Brain Models (LBMs) as a new paradigm in neuroscience, driven by foundation AI models and large-scale data. I will describe our development of SwiFT, a Swin fMRI Transformer, designed to capture the complex spatiotemporal dynamics of fMRI. LBMs have already shown advantages in predicting neonatal development, deco-ding individual brain activation, and classifying conditions such as ADHD and depression. Looking ahead, I will discuss how scaling models, data, and computation could bridge the brain’s non-verbal processes with language models, opening new paths for understanding cognition, emotion, and mental health.
To address the growing mental health crisis, this presentation introduces the ‘Large Brain Model(LBM)’ as a new and powerful paradigm in neuroscience. Moving beyond the predictive limitations of traditional scientific models, this work highlights the vast potential of foun-dation AI models powered by large-scale data and computation.
A core methodology discussed is the end-to-end deep learning model, such as the SwiFT (Swin fMRI Trans-former), which was developed to learn the complex spatiotem-poral dynamics of fMRI data.
These LBMs have demonstrated superior performance over existing models in diverse tasks, including pre-dicting neo-natal development, individual brain activation patterns, and classifying conditions like ADHD and depression.
Ultimately, this talk envisions that the scalability of models, data, and computation will create a break-through in under-standing human cognition, emotion, and mental health by bridging the brain’s non-verbal world with language models.
Exploring New Frontiers in MR Research with Cima.X
In brief
This presentation will delve into the innovative Cima.X high gradient platform, highlighting its transformative potential for MR research. We will explore the clinical and scientific applications that Cima.X enables with its platform technologies, as well as current research activities leveraging this groundbreaking technology.
This presentation will delve into the innovative Cima.X high gradient platform, highlighting its transformative potential for MR research. We will explore both the clinical and scientific applications enabled by Cima.X’s platform technologies, which offer unprecedented opportunities for advancements in the field. Through an in-depth examination of the current research activities leveraging this groundbreaking technology, attendees will gain insights into the diverse and impactful ways Cima.X is shaping the future of MR research.
For instance, recent studies have demonstrated Cima.X’s efficacy in enhancing neuroimaging techniques, leading to more precise mapping of brain functions and improved diagnosis of neurological disorders.
Additionally, Cima.X has been instrumental in advancing cardiac imaging, providing clearer images that assist in the early detection and treatment of heart diseases.
Another notable example is its application in oncology, where Cima.X has enabled more accurate tumor localization and monitoring, thereby improving treatment planning and patient outcomes.
Case studies and real-world examples will be presented to illustrate the practical implications and benefits of utilizing Cima.X in various research scenarios, demonstrating its versatility and effectiveness in pushing the boundaries of what is possible in magnetic resonance imaging.
October 31, 2025 (Friday)
Session II - Bridging Neuroimaging, Decision-Making, and Mental Health
Tracking changes in neural representations over time using intensive fMRI.
In brief
A rapidly emerging approach in human systems neuroscience has focused on acquiring many hours of fMRI data on a few individuals over the course of many months or years, an approach we have termed ‘intensive’ fMRI. In this talk, I will discuss our completed and ongoing efforts to use intensive fMRI to study how neural repre-sentations change over time. First, I will describe our experiments on representational drift, documenting spontaneous, directional changes in neural representations over the course of a year. Second, I will discuss our studies using intensive fMRI to track visual recovery following stroke-related damage to visual cortex.
TBD
The nature and dynamics of working memory representation in the human visual cortex
In brief
Working memory plays a crucial role in various tasks by bridging the gap between sensory input and subsequent cognitive processing. While earlier research shows that working memory representations are present in the early visual cortex, it remains unclear how these coexist with sensory representations of current stimuli without causing conflicts. Using dimension-reduction and population-decoding techniques on fMRI data, I will demon-strate that working memory representations are orthogonal to sensory representations and that their dynamic states accurately reflect the cognitive processes involved in working memory.
TBD
Hierarchical computations underlying the construction of value
In brief
Decision-making theories often assume value signals are retrieved from past experience, but O’Doherty argues that value is actively constructed at choice. He proposes that the brain hierarchically integrates features of potential outcomes, with the lateral orbital and medial prefrontal cortex playing key roles. This process enables flexible, context-sensitive valuation that supports adaptive behavior.
To survive and prosper, humans and other animals need to choose actions leading to beneficial outcomes. Most modern theories of decision-making presume that individuals accomplish this by computing an expected value (or utility) for different decision outcomes, and, all else being equal, committing to the option that yields the highest expected value.
However, a fundamental open question remains: how are these value signals computed in the first place? Attempts to answer this question have predominantly involved an appeal to associative learning whereby a cue or an action acquires value through associations being formed between a hitherto affectively neutral stimulus or action, and an outcome with an extant (perhaps innate) value.
However, past associative history leaves us with an incomplete picture of how value signals for those potential outcomes are computed in the first place. This is because value is not a static variable — instead it can change flexibly and without prior experience depending on both intrinsic and extrinsic factors. Clearly, the brain is capable of flexibly making value-based decisions on the fly based on current motivational and homeostatic states, the context in which a stimulus is being perceived, and the goal that is currently being pursued. Indeed, it is even possible for values to be produced for stimuli that have never before been experienced.
Here we argue that the assignment of subjective value to potential outcomes at the time of decision-making is an active process, in which individual features of a potential outcome of varying degrees of abstraction are represented hierarchically and integrated in a weighted fashion to produce an overall value judgment.
The active construction of value from a weighted combination of underlying features naturally endows the decision-making agent with the capability to: (a) generalize value judgments across stimuli encountered in the environment, even novel ones, provided judgments about the underlying features can be made and (b) flexibly change the weights assigned to attribute features based on changes in internal motivation/homeostasis and/or external context.
We implicate the lateral orbital and medial prefrontal cortex in this function, situating these areas more broadly within a hierarchical integration process that takes place throughout the cortex for the ultimate purpose of valuing options to guide decisions. This mechanism confers on an organism the means to rapidly alter behavior following a sudden change in either internal motivation or the external context or goal, therefore lying at the core of the adaptive control of behavior.
Naturalistic Paradigms and Real Rewards for Unveiling Neurocognitive Mechanisms in Decision-Making and Addiction
In brief
Reinforcement learning and decision-making frameworks have advanced our understanding of human behavior, yet traditional tasks often lack ecological validity. By integrating naturalistic paradigms—such as driving, navigation, and movie-watching—and developing MRI-compatible tools to deliver real-world rewards like nicotine, we aim to address the limitation and reveal neurocomputational mechanisms underlying impulsivity and addiction. These approaches highlight the promise of naturalistic paradigms for capturing individual differences and providing novel insights into the neural basis of addictive behaviors.
The reinforcement learning and decision-making framework has significantly advanced our understanding of the neuro-cognitive processes underlying human behavior and psychopathology. However, traditional lab-oratory paradigms often fail to accurately simulate real-world behaviors and rewards due to their over-simplified contexts and limited eco-logical validity. To overcome these limitations, my lab has integrate naturalistic paradigms and natural rewards directly into neuroimaging studies.
In one line of research, we employed naturalistic tasks such as real-time driving, real-time navigation, or movie-watching paradigms combined with neuroimaging and computational approaches to elucidate in-dividual differences in impulsivity and addiction.
In another series of studies, we utilized an MRI-compatible vaping device to directly investigate neural pro-cessing of primary drug rewards in regular smokers.
Collectively, these studies exemplify the potential of naturalistic paradigms and advanced neurocomputational approaches to simulate real-world situations and characterize individual differences, offering novel insights into addiction.
Dopamine and glutamate hypotheses for treatment response in schizophrenia
In brief
Treatment-resistant schizophrenia (TRS) affects one-third of patients and may reflect a distinct biotype marked by medial frontal glutamatergic dysregulation rather than uniformly elevated presynaptic dopamine. Evidence from spectroscopy, molecular imaging, and neuromelanin-sensitive MRI indicates multivariate deep phenotyping is needed to stratify patients and guide clozapine and glutamate-targeted interventions.
Roughly one-third of individuals with schizophrenia meet criteria for treatment-resistant schizophrenia (TRS), underscoring the need to clarify biological correlates of treatment response. This presentation aims to synthesize convergent evidence across spectroscopy, molecular imaging, and neuromelanin-sensitive MRI to evaluate the dopamine and glutamate hypo-theses of treatment response, with emphasis on TRS and clozapine resistance.
TRS may represent a biotype characterized by medial frontal glutamatergic dysregulation rather than uniformly heightened presynaptic dopamine function. While dopaminergic and glutamatergic markers are abnormal at the illness level, they in-completely account for treatment response.
Deep phenotyping with multivariate approaches is warranted to parse heterogeneity and to integrate glutamatergic and dopaminergic mechanisms—particularly when stratifying candidates for clozapine and next-generation glutamate-targeted interventions.
Neuroimaging fingerprint for treatment-resistant schizophrenia
In brief
About one-third of patients with schizophrenia are classified with treatment-resistant schizophrenia (TRS), which potentially involves a distinct pathophysiology compared to its non-TRS counterpart. This talk will highlight the potential of magnetic resonance imaging to probe the neural mechanisms underlying TRS and discuss key challenges for advancing its study.
Schizophrenia is a chronic, severe mental illness characterized by psychotic symptoms such as hallucinations and delusions, often coupled with cognitive and social impairments. Dopamine D2 receptor antagonists and partial agonists are the mainstay of treatment of schizophrenia. However, approximately thirty percent of patients with schizophrenia do not respond to first-line antipsychotic treatments, a condition referred to as treatment-resistant schizophrenia (TRS). Notably, the majority of patients classified with TRS fail to show any response to antipsychotics from the onset of their treatment, leading to the implication that TRS potentially involves a distinct pathophysiology compared to its non-TRS counterpart.
The introduction of neuroimaging techniques enabled the investigation into the living brain of patients with schizoph-renia. Advances in these methods have provided substantial evidence regarding the pathophysiology of schizophrenia, parti-cularly the dysregulated dopamine system. However, research on TRS remains in its early stages. In this context, I’m going to discuss both the potential applications of magnetic resonance imaging and the challenges that must be add-ressed in future studies of the pathophysiology of TRS.
