Speakers/Tutorial leaders

Speakers

Chung-ke Chang, Ph.D.

Chief Strategy Officer

Taiwan Biobank, Academia Sinica, Taiwan

Balancing Open and Closed Data Governance in Biobanking: Finding the Right Formula

We now live in a world where medical breakthroughs depend on vast datasets, but strict privacy laws limit what can be shared. Data governance is critical for balancing data accessibility with ethical and legal responsibilities. The Taiwan Biobank, a national research infrastructure, houses over four million biological samples and 1.5 petabytes of data from 200,000 participants. While researchers can access these resources through formal application processes, the data are classified as personal data under Taiwanese law, restricting full open access. However, the Taiwan Biobank also generates summary statistics and usage data, which are openly available through a public portal. This dual nature of data—some restricted, some openly shared—necessitates a hybrid governance approach, following the principle of being "as open as possible, as closed as necessary." In this talk, I will explore how the Taiwan Biobank integrates elements of both open and closed governance models, the challenges this presents, and why a nuanced approach is essential for balancing scientific progress, privacy protection, and public trust. As data-driven research expands, this discussion is critical—not just for biobanking but for all large-scale biomedical and health research initiatives worldwide.

Annabel Chen, Ph.D.

Professor

Centre for Research and Development in Learning, Nanyang Technological University, Singapore

Building Trustworthy and Transparent Neuroscience Research: The NTU Singapore Experience

Open science is reshaping how neuroscience research is conducted, shared, and evaluated. This talk will explore how institutions like Nanyang Technological University (NTU) have developed robust open research infrastructures—ranging from policies and repositories to cultural practices—to support transparency, reproducibility, and responsible data sharing. Highlighting NTU's journey and milestones, including CoreTrustSeal certification, FAIR data practices, and cross-disciplinary collaborations, we will discuss how neuroscience researchers can adopt open science principles to enhance scientific integrity and foster global collaborations. Practical examples will demonstrate how open data, metadata standards, and community engagement can be leveraged to accelerate discovery and ensure long-term research value in neuroscience.

Erin W. Dickie, Ph.D

Assistant Professor

Department of Psychiatry, University of Toronto, Canada

Erin W. Dickie, Ph.D.

Staff Scientist, Krembil Centre for Neuroinformatics, Centre for Addiction and Mental Health; Assistant Professor, Department of Psychiatry, University of Toronto

Day 1 (Applications): Leveraging Neuroimaging to Uncover Heterogeneity in Mental Illness: Insights from Transdiagnostic Samples

85% of people experience mental illness. However, mental illness can follow many different paths. Having one disorder increases the risk of developing others. Neuroimaging can offer unique insights into the structure and function of the brain and, thus, offers hope of identifying biomarkers at the highest risk for poor outcomes, a necessary step for tailoring care to mitigate this risk. Present research can have a greater translational impact by looking beyond case-control comparisons that assume some uniformity within the patient population to examine clinically meaningful sources of heterogeneity. We will present recent work that uses neuroimaging to model sources of cognitive function in participants with schizophrenia spectrum disorders, areas of untreated clinical need. We will also present available data and introduce new cohort studies that are following children and youth over time to look at the link between brain development and trajectories of mental health, especially those imaging youth with multiple mental health diagnoses.

Day 2 (Open Tools - ciftify): Reproducible Approaches for Cortical Surface-Based MRI Data Analysis

We will discuss the benefits of standardized, reproducible workflows in neuroimaging research that account for individual variability in brain anatomy. Specifically, we will explore the advantages of a surface-based approach using the CIFTI file format. We will introduce reproducible, open software toolkits that will allow participants to integrate this approach into their own analysis. At the end of this presentation, students will learn about 1) reproducible preprocessing workflows (BIDS apps) that generate CIFTI files 2) methods for analyzing these files in Python and R, and 3) visualizing CIFTI files using Python and the Connectome Workbench.

Peer Herholz, Ph.D.

Assistant Professor, School of Communication, Northwestern University, USA

Research Affiliate, Vision and Computational Cognition Group, Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig, Germany

How open and reproducible science has supported my cognitive and computational neuroscience

Open and reproducible science has been the very bedrock of my work in cognitive and computational neuroscience, providing the data, tools, and teaching materials that made my research possible. Thanks to openly shared and standardized neuroimaging and behavioral datasets, I was able to run a multitude of studies across a diverse set of questions, experimental designs and populations, as well as enabling a focus on generalization and replicability. The respective analyses were made possible by adopting and contributing to open‑source analysis toolkits, ranging from univariate statistics over connectivity to complex deep

learning analyses. Leveraging open educational resources, from freely available lecture slide decks to hands‑on workshops, has not only tremendously amplified my skillset and knowledge but also my ability to train students and colleagues worldwide in cutting‑edge methods. In this presentation, I’ll walk through the respective resources, outlining the vast open and reproducible neuroscience ecosystem, to show how they can be adopted to drive research.

Wei Jeng, Ph.D.

Associate professor, Department of Library and Information, National Taiwan University

Towards Better Science: Open Science in a Data-Intensive Perspective

Dr. Wei Jeng explores the contemporary significance of open science practices in today's data-rich research environment. The presentation highlights key components of the notion of open science: Open research data sharing, reproducibility practices including preregistration to combat questionable research practices (QRPs), and citizen science initiatives that expand participation in scientific discovery. The talk provides a comprehensive overview of the current landscape of open science, addressing challenges faced by researchers in implementing these practices and identifying emerging opportunities that could shape the future of scientific inquiry.

Jean-Baptiste Poline, Ph.D.

Professor

McConnell Brain Imaging Centre, Neurology and Neurosurgery, School of Computer Science, McGill University, Canada

What type of data and computing infrastructure do we need for developing Robust Imaging biomarkers ?

Research in neuroscience and neuroimaging is very likely inefficient. One of the factors is that it is often seen as a personal endevour. In this presentation, I will lay out some of the neuroscience and neuroimaging research challenges, and in particular how do we the challenges of reproducible and generalizable studies in the context of data distributed with various legal and ethical constraints. While there are many technical challenges, the academic incentive system and the sociological aspect of research are playing a key role in shaping our developments. In this presentation I will take a few examples of open neuroinformatics projects developed in my laboratory and in collaboration with others to help data sharing and data analysis designed to help the development of some imaging biomarkers for Parkinson Disease. I will open the discussion on some key aspect to consider, for instance how these open infrastructures can be maintained, sustained, and how their governance structure should be thought to balance their longevity and agility.

Jessica Turner, Ph.D.

Professor

Department of Psychiatry and Behavioral Health, Wexner Medical Center, The Ohio State University, USA

How many degrees of freedom does a dataset have?

Neuroimaging data are expensive to collect, and valuable beyond the purposes of the original study. "Open Science" as applied to neuroimaging studies can refer to data collection techniques, processing and analysis pipelines, and of course, the data itself. The past 25 years have seen a number of approaches to neuroimaging datasharing, all with the goal of driving toward more open, rigorous, and innovative science. The ENIGMA consortia over the past 12 years have spearheaded multiple approaches to data sharing, taking into account the various and evolving constraints placed on neuroimaging datasets by their protocols, sponsors, institutions, and national policies. Focusing on the various projects in the ENIGMA Schizophrenia working group, I will present a number of options that have been practically applied under different constraints.

Elizaveta Baranova, Ph.D.

Researcher

Graduate Institute of Mind, Brain and Consciousness, Taipei Medical University, Taiwan

A student's perspective on learning and adopting open science practices

Open science provides numerous tools to promote transparent and reproducible research. In this talk, I will introduce my own path of discovering and implementing different tools, along with the corresponding obstacles that were encountered. I will firstly discuss my experience of resources available to students who want to learn about open science and follow its principles in their work. I will then reflect on my experience of going on to promote open science within a local student community. Finally, I will talk about experiences of using open data in my work. In each case, I will seek to highlight the advantages I experienced along with the barriers faced. Of course, other students encounter different issues, but some of my experiences are shared and may help ease the transition for future students, especially if these obstacles are openly addressed and, ideally, solved.

Anthony Edward Galassi

Computer Programer

National Institute of Mental Health, Brain Imaging Data Structure, & OpenNeuroPET

The Brain Imaging Data Structure (BIDS) Autobiographically: An open source maintainers tale

Attendees will get to follow the journey of Anthony Galassi and the BIDS Standard as they were both introduced into the world of Neuroscience over the course of a decade. This talk serves to provide attendees with the origin stories of both BIDS and one of its current maintainers. Through an auto-biographical lens, they'll get to see how an utterance or two can not only spiral into a career in Open Science, but also lead to the building of a community with hundreds of active contributors, the creation of a ridiculous amount of software, and the growth from a convention for naming MR files to an over 500 page technical specification.

Tutorial leaders

Chun-Hsien Hsu, Ph.D.

Assistant Professor

Institute of Cognitive Neuroscience, National Central University

Open Science MEG/EEG Advanced Tools

This tutorial will provide a basic MEG/EEG analysis background and explore advanced analysis tools for EEG/MEG open science projects. In particular, we will focus on MNE-Python, one of the main free analysis programs available, and practices for reproducible research. This lecture is designed for students and researchers who are either beginning or planning to launch their own EEG/MEG studies. We will attempt to use the knowledge obtained from this tutorial lecture further in one of the third-day hackathon projects.

The lecture is divided into four key sections:

A. Basics of EEG/MEG Signals and Their Applications – Provide an introduction to the fundamentals of EEG and MEG, including their neural basis, recording techniques, and some important pre-processing steps.

B. Introduction to MNE-Python – A hands-on introduction to MNE-Python, a widely used open-source toolbox for EEG/MEG analysis, covering its core functionalities and applications in neural data processing.

C. Building Reproducible Analysis Pipelines – Practices for structuring analysis workflows in Python, ensuring transparency, efficiency, and reproducibility in research.

D. Exploring Open Datasets and Tools – An overview of freely available EEG/MEG datasets, demonstrating how to integrate them into your own projects for secondary analyses and collaborative research. We will also introduce some additional available analysis tools that can be used in your open science projects.

Through this lecture, participants will gain practical knowledge of open-source tools and methodologies that facilitate high-quality, reproducible, open-science EEG/MEG research. Whether you are new to EEG/MEG analysis or looking to refine your workflow, this session will provide valuable insights into open science applications in neuroimaging.

Stefano Moia, Ph.D.

Researcher

General Chair, OHBM Open Science Speical Interest Group

Department of Cognitive Neuroscience, Maastricht University, The Netherlands

Building open scientific softwares and communities: the case of physiopy

Open scientific tools are the foundations of open science, the instruments through which we can build our experimental pipelines, from data collection and curation to analysis and reporting. They don’t come for free, the cost being either in trust or engagement, understanding, and time, but such investment can become a benefit in our day-to-day job. This lecture will introduce you to the fundamentals of open scientific development, both of software and of communities, from version control systems to automation, from licences to scientific engagement. We will look at the physiopy community as a case example, with the aim to improve your familiarity with open tools, improve trust and/or engagement, lower risks in adopting such tools, or at the very least provide a framework that can become beneficial to your own pipeline development.

Michelle Wang

Researcher

Department of Quantitative Life Sciences, McGill University, Canada

Ju-Chi Yu, Ph.D.

Researcher

Open Science Room Chair, OHBM Open Science Speical Interest Group

Campbell Family Mental Health Research Institute, The Centre for Addiction and Mental Health, Canada

Yaron Caspi Ph.D.

Researcher

Department of Psychology, National Taiwan University, Taiwan

Open Science MEG/EEG Advanced Tools

This tutorial will provide a basic MEG/EEG analysis background and explore advanced analysis tools for EEG/MEG open science projects. In particular, we will focus on MNE-Python, one of the main free analysis programs available, and practices for reproducible research. This lecture is designed for students and researchers who are either beginning or planning to launch their own EEG/MEG studies. We will attempt to use the knowledge obtained from this tutorial lecture further in one of the third-day hackathon projects.

The lecture is divided into four key sections:

A. Basics of EEG/MEG Signals and Their Applications – Provide an introduction to the fundamentals of EEG and MEG, including their neural basis, recording techniques, and some important pre-processing steps.

B. Introduction to MNE-Python – A hands-on introduction to MNE-Python, a widely used open-source toolbox for EEG/MEG analysis, covering its core functionalities and applications in neural data processing.

C. Building Reproducible Analysis Pipelines – Practices for structuring analysis workflows in Python, ensuring transparency, efficiency, and reproducibility in research.

D. Exploring Open Datasets and Tools – An overview of freely available EEG/MEG datasets, demonstrating how to integrate them into your own projects for secondary analyses and collaborative research. We will also introduce some additional available analysis tools that can be used in your open science projects.

Through this lecture, participants will gain practical knowledge of open-source tools and methodologies that facilitate high-quality, reproducible, open-science EEG/MEG research. Whether you are new to EEG/MEG analysis or looking to refine your workflow, this session will provide valuable insights into open science applications in neuroimaging.

主辦單位 Organiser：台北醫學大學心智意識與腦科學研究所 Graduate Institute of Mind, Brain and Consciousness (GIMBC), Taipei Medical University

贊助單位 Sponsor：國家科學及技術委員會 National Science and Technology Council

In collaboration with the Open Science Room (OSR)🔗 of the Open Science Special Interest Group (OS-SIG)🔗 of the Organization of Human Brain Mapping (OHBM)