Spatial analysis of gene expression is an essential tool for comprehensive studies of complex, highly heterogeneous tissues, such as the brain, tumors, and developing organs. RNAscope HiPlex v2 assay is an advanced, in situ hybridization (ISH) tool to understand vital gene expression patterns at single cell resolution in highly complex tissues. The assay generates precise gene expression data, enabling comprehensive spatial studies from 12 targets in FFPE samples and up to 48 targets in Fresh and Fixed Frozen sample types. 

The G-protein coupled receptors (GPCR) are an important component of pain modulation and are one of the most important therapeutic targets in pain medicine. However, detection of GPCRs can be challenging due to difficulties in obtaining suitable antigen accessibility and their low expression levels. In this webinar, using the highly sensitive HiPlex v2 assay with the upcoming RNAscope HiPlex v2 Whole Slide image registration software, we will demonstrate spatial mapping of G-protein coupled receptors (GPCR) within the normal mouse brain at single cell resolution. We further demonstrate detection of GPCRs and Neuropeptide Y (NYP), known to be involved in pain perception and transmission in the brain, using the integrated RNA-protein codetection workflow. 

The study provides deeper insights into understanding the spatial crosstalk as well as functional significance of different cell populations within the various brain regions thereby leading to broader understanding of disease pathology.

Learning Objectives:

1. Discuss the latest development in the HiPlex v2 assay including image registration software for WSI registration and downstream image analysis.

2. Explain spatial gene expression mapping for visualizing GPCR family receptor, within the normal mouse brain at single cell resolution. 

3. Identify cell type specific gene expression of opioid receptors in different regions of the brain to predict functional significance of different cell types in the pain pathway.
 

Deep–Brain Stimulation (DBS) has become an effective means of treating a variety of neurological conditions. Most DBS systems work are "open-loop," and apply constant stimulation over the lifetime of the treatment. The next generation of DBS systems are "closed-loop,” recording neural activity and adjusting stimulation to adapt to the user’s needs. Different ways of controlling DBS, however, will likely produce different experiences of agency for their users—with differing moral outcomes. In this talk, I suggest that the constituent components of a closed-loop system change the experience of user agency in ways that could call the trustworthiness of these devices, and the institutions that create them, into question.

Learning Objectives:

1. Discuss components of an adaptive brain computer-interface.

2. Identify the ethical issues that arise from each of the component tasks performed by brain computer interfaces.

3. Evaluate the design and user-experience of brain-computer interfaces to anticipate and address ethical issues.

Artificial neural networks can be useful for studying brain functions. In cognitive neuroscience, recurrent neural networks are often used to model cognitive functions. I will first offer my opinion on what is missing in the classical use of recurrent neural networks. Then I will discuss several lines of ongoing efforts in our group to move beyond the classical recurrent neural networks by studying multi-system neural networks (the talk will focus on two-system networks). These are networks that combine modules for several neural systems, such as vision, audition, prefrontal, hippocampal systems. I will showcase how multi-system networks can potentially be constrained by experimental data in fundamental ways and at scale. I will demonstrate how we can gain new neuroscientific insights by studying these multi-system models.

Learning Objectives:

1. Discuss the challenge of classical recurrent neural network approach to cognitive neuroscience.

2. Discuss multi-system neural network models and their applications to neuroscience.

3. Discuss how multi-system models can shed light on fundamental questions such as the nature of working memory limit.

Animal models present a unique opportunity to understand molecular mechanisms of pathological changes involved in cognitive dysfunction in neurological disorders. However, most of the work focusing on cognition in mice have yet to translate to new effective drugs in humans. First generation mouse models of Alzheimer’s disease have focused on transgenic expression of human genes, usually driven by neuronal promoters. Although these models have been informative to understand pathology, they failed to reproduce the subtle age and brain region relevant pathology observed in humans. Moreover, most of the analysis in this first-generation mouse models suffers from lack of translational approaches to investigate cognition and mesoscopic brain changes. Next-generation mouse models provide opportunities to understand the role of humanized genes in pathology and cognitive function. We are developing a fully translational platform for testing how these interactions manifest as cognitive and molecular changes and how translatable these are to human data. Touchscreen tests that are similar, and sometimes identical to human tests improve translational of cognitive function testing in health and disease. In parallel, we are optimizing in vivo imaging approaches routinely used in humans for parallel studies in mice. Given the automated and standardized approach of this platform, data can be organized and deposited in repositories (such as www.mousebytes.ca), facilitating open science goals. I will focus on our collaborative efforts to study the contribution misfolded and toxic alpha-synuclein for cognitive dysfunction and how risk factors in dementia affect cognition. I will also explore potential molecular mechanisms that can be used to develop new therapeutics in open translational science. This work provides a road map on how to improve reproducibility and predictive power of preclinical data to impact clinical outcomes. 

Learning Objectives:

1. Discuss approaches to integrate open science in pre-clinical research.

2. Discuss toxicity of misfolded proteins.

3. Discuss the principles of using high level cognitive tests to study dementia in animal models.

Mild traumatic brain injuries (mTBIs) are a prevalent health issue in North America. There is increasing pressure to utilize ecologically valid models of closed-head mTBI in the preclinical setting to increase translatability to the clinical population. The awake closed-head injury (ACHI) model described here produces clinically relevant behavioral deficits in rodents without the need for a craniotomy or the use of an anesthetic. The technique involves rapidly accelerating the animals head, rather than striking it, so it does not normally induce fatalities, skull fractures, or brain bleeds, and is more consistent with being a mild injury. Indeed, the mild nature of the ACHI procedure makes it ideal for studies investigating repetitive mTBI (r-mTBI). Our recent work indicates that the ACHI is associated with a distributed brain injury that results in microbleeds, microglia activation and synaptic pruning. This in turn results in changes in synaptic plasticity, the main biological model for learning and memory processes. Future studies will continue to explore deficits in synaptic plasticity, as well as how an aged brain is affected by incurring an mTBI early in life.

Learning Objectives:

1. Discuss the merits and pitfalls of The Awake Closed Head Injury (ACHI) model for brain injury research. 

2. Describe how mild traumatic brain injury can affect the structure and function of the brain.  

3. Discuss opportunities for future studies using a clinically relevant model of mild traumatic brain injury like the ACHI.

Explainable AI (XAI) aims at helping you understand and interpret predictions made by AI techniques and tools. Recently, there has been a growing interest in XAI as many users want explanations to the results produced by AI models. For deep neural networks, saliency of neurons defines which features learned by neurons are most salient for the predictions. This allows us to recognize and understand which inputs are attributed to model performance. In this talk, I present an approach to explainable deep neural networks by using saliency backpropagation which propagates saliencies of nodes at the output layer to inputs. I highlight existing saliency back propagation methods and their applications to biomarker discovery from functional MRI (fMRI) data and to drug mechanism prediction.

Learning Objectives:

1. Discuss the importance of Explainable AI (XAI) and the approaches to XAI.

2. Explain the method of saliency backpropagation for XAI.

3. Describe the applications of XAI for drug discovery.

4. Discuss biomarker discovery from functional MRI analysis as an XAI application.

Black Americans have an increased risk for age-related cognitive decline and Alzheimer's disease (AD) compared to other ethnic and racialized groups. However, risk factors for age-related cognitive decline and AD have been defined based largely on data from white populations. This seminar will explore how a few of these "well-established" risk factors impact Black individuals differently than white individuals. This and other analogous research continue to show the importance of understanding risk factors within the Black population, not in comparison to the white population, as a best practice toward health equity research.

Learning Objectives:

1. Identify disparities for risk factors associated with brain health specific to the Black community.

2. Discuss how genetic and neuroimaging risk factors impact brain health and cognition.

3. Identify gaps to address in order to minimize cultural bias in recruitment for clinical trials and cohort studies.

In this talk, I will discuss the development of interpretable machine learning models to test scientific hypotheses, with a specific focus on spinal motor control. Voluntary movement requires communication from cortex to the spinal cord, where a dedicated pool of motor units (MUs) activates each muscle. The canonical description of MU function states that cortex cannot flexibly control MUs independently, but rigidly supplies each pool with a single common drive. Here, we test this “rigid control” theory through both 1) novel experiments in which many MUs were stably recorded while macaques generated widely varying forces, and 2) the development of a latent variable model that represents the rigid control theory. By testing whether this latent variable model could accurately fit the recorded MU data, we were able to quantitatively test the rigid control theory. We found that a single latent variable is unable to describe MU activity across widely ranging behaviors, demonstrating that, contrary to canonical beliefs, MU activity is flexibly controlled to meet task demands. 
 
Learning Objectives:

1. Describe latent variable models.

2. Discuss theories of motor unit control.

3. Discuss fitting machine learning models to data to test scientific hypotheses.

Natural Language Processing is experiencing a phenomenal capabilities acceleration since the introduction of Large Language Models (LLMs). This success stems in part from the ability to ingest diverse and vast amounts of unstructured text to quantify normative, unlabeled language. It is not obvious however how to make use of LLMs to analyze language produced in the context of psychiatric evaluation and intervention, and how to incorporate decades-worth of neuropsychiatric knowledge in the form of questionnaires, inventories, constructs, manuals, papers, etc., not to mention the personal experience of psychiatrists and patients. We will discuss recent efforts to leverage this knowledge in several applications, including analysis of interviews for differential diagnosis of psychosis and its risk states, discrimination of fibromyalgia and neuropathic pain, prediction of treatment outcomes in cocaine addiction, and inference of progression in psychotherapy. We will also discuss similar efforts in modeling multimodal data streams, e.g., the combination of qualitative assessments and neuroimaging in the clinic with at-home digital data in ongoing clinical trials.

Learning Objectives:

1.  Identify how current language analytics can be applied in different mental health assessment tasks.

2. Generalize the concept of Interpretable AI to models arising from different data modalities and analytic tasks.

3. Evaluate the advantages and drawbacks of multimodal classification/regression models.

Children efficiently develop their visual systems through learning from their environment. How this development unfolds in noisy real-world data streams remains largely unknown. Deep neural networks trained on large annotated visual datasets have become state-of-the-art models for both visual recognition tasks and predicting neuronal responses in the primate visual stream. However, the large number of annotations required to train these networks made them implausible as models of real visual development and learning. In this talk, I will first show that the recent rapid progress in unsupervised learning has largely closed this gap. We find that neural network models learned with deep unsupervised contrastive embedding methods achieve neural prediction accuracy in multiple ventral visual cortical areas that equals or exceeds that of the models supervised on the labels, even when trained on noisy and limited first-person view datasets collected from infants. I will then show that augmenting these models with memory enables capturing human learning dynamics in both real-time and life-long time scales and that the more recent unsupervised learning algorithms diverge from being human-like compared to earlier algorithms.

Learning Objectives:

1. Explain why categorization-trained deep neural networks cannot model how humans develop their visual system.

2. Describe how contrastive learning algorithms train the neural network models from images at a high level.

3. Paraphrase how to model the real-time and life-long visual learning dynamics using the unsupervised learning algorithms.
 

Alzheimer’s and Parkinson’s diseases (AD, PD) are the two most common neurodegenerative disorders. Both diseases are characterized by abnormal aggregates in the brain, which include alpha-synuclein, tau and amyloid beta (Aβ). These proteins share a common pathological mechanism whereby aggregation leads to neuronal toxicity and plaque formation. A key goal in advancing AD/PD drug discovery is establishing disease models that effectively mimic pathology observed in vivo. We have developed consistent, scalable methods to generate active neurotoxic protein aggregates in vitro. Recently we generated fibrils of pyroglutamate Aβ, an N-terminally truncated peptide modified by glutaminyl cyclase present in late-stage AD, and fibrils of the tau dGAE fragment, which we show more closely demonstrates AD-like fibril pathology than other tau fibrils. Finally, we have also generated co-polymer alpha-synuclein and tau fibrils. These novel molecular tools expand the range of constructs available to study neurodegenrative disease and facilitate AD/PD drug discovery efforts.

Learning Objectives:

1. Discuss various pathogenic protein aggregates involved in Parkinson’s and Alzheimer’s disease.

2. Explain how aggregates generated in the laboratory mimic those isolated from diseased brains and can be used to study neurodegeneration.

3. Discuss novel aggregates being targeted at the forefront of therapeutic development.

Neural networks are a powerful tool for modeling neural activity in the brain. In this talk, I will discuss how these models have helped in my own research and highlight recent work building neural networks that not only reproduce observed neural activity but also shed light on the underlying circuit mechanisms as they evolve a brain-like connectome through training. To more fully understand the brain, however, we need to take these modeling efforts beyond single circuits and tasks and consider the complex interactions between multiple systems: perception, cognition, action, memory, and many others. Achieving this goal requires models that bridge these systems, but this presents several challenges, including navigating the vast design space of potential models. To address this issue, I will describe ongoing work where we are developing a high-throughput pipeline to systematically build and evaluate recurrent neural network models on multiple datasets. By doing so, we hope to extract general design principles, or identify combinations of tasks, datasets, and models that are not currently well-explained, allowing us to continue to refine our understanding of both biological and artificial neural networks.

Learning Objectives:

1. Describe recurrent neural network (RNN) models.

2. Discuss ways in which RNNs are used to study the brain.

3. Evaluate the advantages and drawbacks of using RNNs as models for understanding the brain.

Selective vulnerability of brain regions and specific neural populations in neurodegeneration remains little understood. We hypothesize that pre-existing physiological differences in neurons and in surrounding glia across brain regions are likely to contribute to their selective vulnerability in disease. Spinocerebellar Ataxia type 1 (SCA1) is a dominantly inherited neurodegenerative disease caused by an abnormal expansion of polyglutamine (polyQ) repeats in the ATXN1 protein and characterized by early and severe cerebellar degeneration. Using histological approaches, we first demonstrated that the cerebellum of SCA1 mice does not undergo degeneration uniformly with worse Purkinje cell atrophy and gliosis in the posterior vermis, as seen in patients with SCA1. Using our novel region-specific dissection method and RNA sequencing, we identified genes and pathways across these three cerebellar cortical regions in wild-type and in SCA1 mice that may underlie different vulnerability. Finally, using fiber photometry we examined whether these histological and molecular pathologies result in different extent of neuronal dysfunctions in SCA1. This knowledge will increase our understanding of intraregional cellular, molecular and functional Purkinje cell differences in physiology and pathology.

Learning Objectives:

1. Discuss how inter- and intraregional heterogeneity of neurons and glia is a key characteristic of a healthy brain, while selective regional neuronal vulnerability and gliosis is a key feature of neurodegenerative diseases.

2. Discuss how molecular differences underlying inter-and intraregional heterogeneity of the brain likely contribute to selective vulnerability in brain disease.

3. Explain how larger cellular and molecular changes correlate with neuronal dysfunction in vulnerable regions, indicating that there is a threshold for dysfunction and/or compensatory changes in disease resistant regions.

Disability activists have sometimes claimed their disability has actually increased their well-being. Some even say they would reject a cure to keep these gains. Yet, these same activists often simultaneously propose improvements to the quality and accessibility of assistive technology. However, for any argument favoring assistive over curative technology (or vice versa) to work, there must be a coherent distinction between the two. This line is already vague and will become even less clear with the emergence of novel technologies. This paper asks and tries to answer the question: what is it about the paradigmatic examples of curative and assistive technologies that make them paradigmatic and how can these defining features help us clarify the hard cases? This analysis will begin with an argument that, while the common views of this distinction adequately explain the paradigmatic cases, they fail to accurately pick out the relevant features of those technologies that make them paradigmatic and to provide adequate guidance for parsing the hard cases. Instead, it will be claimed that these categories of curative or assistive technologies are defined by the role the technologies play in establishing a person’s relational narrative identity as a member of one of two social groups: disabled people or non-disabled people.

Learning Objectives:

1. Discuss why it is important for technologists to clearly distinguish between assistive and curative neurotechnology.

2. Describe, evaluate, and compare each of the major theoretical positions that offer definitions for assistive technology.

3. Describe and justify your own theoretical position regarding how you should define assistive technology and distinguish it from curative technology. 

Human induced pluripotent stem cells (iPSCs) are immortal cell lines that can be derived from skin or blood of any individual. Their unique property of pluripotency means they can be differentiated into any cell type in dish. A major challenge in working with iPSCs is to reliably and efficiently differentiate them into desired cell types, such as midbrain dopaminergic (mDA) neurons. We have adapted protocols to convert iPSCs into mDA neurons by taking lessons from early embryo development. Although most iPSC lines can produce mDA neurons, there is considerably differences in the efficiency of mDA differentiation when cell lines are compared. The most sensitive variable of the protocol is the strength of WNT signal activation.  We have developed two sets of tools to address the variable differentiation of iPSC lines. The first is non-invasive monitoring by quantification of mDA lineage-specific secreted factors, such as Trefoil Factor 3. The second technology is optimized cryopreservation methods of mDA neural progenitor cells. Cells are differentiated beyond the time-window of WNT activation prior to freezing them. Both technologies have significantly increased the reliability and predictability of mDA differentiation. The cryopreserved poised neural progenitors have also facilitated collaborations with laboratories with diverse expertise. Non-invasive, real-time monitoring of differentiation and cryopreservation of ready-to-go progenitor cells will also aid the manufacture of clinical-grade neural progenitor cells for regenerative transplantation therapies.

Learning Objectives:

1. Describe the signalling pathways that are manipulated during the first 16 days of iPSC differentiation into dopaminergic progenitor cells.

2. Describe how small changes in WNT signal activation by CHIR99021 can affect dopaminergic differentiation.

3. Describe the experimental evidence that shows endogenous WNT signalling is contributing to the dopaminergic differentiation process.

Some of the earliest symptoms of Alzheimer’s disease (AD) are loss of orientation, wandering and misplacing items. One of the first regions vulnerable to AD pathology is the entorhinal cortex (EC), and together with hippocampus (HPC), the EC-HPC circuit is intimately involved in several memory functions, including spatial memory. Using mouse models of Alzheimer's disease, I will show how amyloid beta and tau pathologies affect neuronal and network function in the EC-HPC circuit. More specifically, I will show how the neural correlates of spatial memory, the place cell and the grid cells are impacted by AD pathologies. Further, I will demonstrate how neuronal activity influences pathology and if modulating the neuronal activity can rescue neurodegeneration. 

Learning Objectives:

1. List which regions in the brain are vulnerable to Alzheimer’s disease pathology.

2. Discuss how amyloid beta and tau pathologies impact spatial memory.

3. Explain how manipulation of neuronal firing can improve or worsen pathology.