Speakers

Current and future scenarios for open scientific publishing and reviewing

Abstract: Communicating our research is a fundamental part of our work as scientists, and a duty towards the society. Typically this communication happens through articles published in scientific journals, after having been reviewed by our peers. We will discuss several aspects of scholarly communication and in particular of scientific publishing and peer review. Some of these aspects are (un-)surprisingly in clear contrast with the idea of science as an open and collaborative public mission. Several solutions and improvements have been proposed and sometimes implemented over the years, on some of them there’s wide consensus, on others there is not. And even when there is consensus to change, the change is slow. We will explore the state of the art and different future perspectives, and discuss our experiences and expectations.

Bio: Daniele is an associate professor in the department of Data Analysis of the Faculty of Psychology and Educational Sciences at Ghent University. His team focuses on methodological and computational aspects of neuroscience research, and on the dynamical networks subserving function, as well as thorough statistical validation of the results. They develop new techniques for inferring connectivity architectures from the dynamics of the recorded data, in challenging cases of short, noisy and redundant time series, as those encountered in neuroimaging. Daniele cares about open science and ways to improve the review/editorial process. He is an editor at several journals in his field, including PLOS Computational Biology, PLOS One, NeuroImage, Brain Topography, Network Neuroscience. Visit Daniele’s Publons review profile for more.

How I Fail in Open Science

Abstract: I am an open science enthusiast in the same way as I am runner. I want to be great at it and want to take all the steps, BUT life gets in the way, and I only do it here and there. With both things, I’ve come to the conclusion that doing something is doing better than nothing. I’ve also found that sharing my progress (or lack of it) is better than hiding it, as it helps for me to stay motivated. In the talk I will share some of my experiences with open science - trying, but mostly failing to do it. The examples will be focused on preprints and sharing code and data. I will discuss incentives in academia, which can be frequently blamed for avoiding open science. Finally I will then discuss some strategies everybody can use to avoid failing at open science in the future.

Bio: Veronika is an assistant professor at the Medical Image Analysis group, Eindhoven University of Technology since February 2017. She received her Ph.D. from the Delft University of Technology for her thesis Dissimilarity-Based Multiple Instance Learning in 2015. As part of her PhD, she was a visiting researcher at the Max Planck Institute for Intelligent Systems in Tuebingen, Germany. From 2015 to 2016 she was a postdoc at the Biomedical Imaging Group Rotterdam, Erasmus MC. Her research interests are centered around learning scenarios where few labels are available, such as multiple instance learning, transfer learning, and crowdsourcing. Next to research, Veronika blogs about academic life. Visit Veronika’s website.

Sharing is caring: how mentoring and open science come together to equalize chances for career development in academia

Abstract: In the spirit of Open Science, we tend to increase an amount of shared resources, such as codes and datasets, nowadays. However, what else can and should we share in academia? In this talk, we’ll discuss mentoring as one example of a human asset which also can and should be distributed. It is no longer the case that mentoring can only be received from your direct supervisor; there are multiple other options, and you can actively search out for mentoring advice from researchers at every career stage and independent from geographical location. Within the Organization for Human Brain Mapping Student and Postdoc Special Interest Group, we are interested in developing new tools assisting early career researchers in accelerating their careers. One of our main initiatives is the International Online Mentoring Programme, which, to date, attracted a few hundreds of OHBM members. After two finished cycles of the programme, we are proud to say we feature multiple success stories: mentees from all around the world getting inspired, finding new jobs and personal grants, and coming back to the Programme as mentors. Additionally, this talk aims to ignite discussion about mentorship as a competence which should be featured as part of a standard research CV and acknowledged on equal terms with other research-related competences i.e. publication record, teaching experience or received grants and awards.

Bio: Natalia Bielczyk is now completing her thesis within the Donders Graduate School, Donders Institute for Brain, Cognition and Behavior, Nijmegen, the Netherlands. From scientific point of view, she is interested in developing new method for data analysis in connectomics, especially in the domain of methods for functional and effective connectivity. She is also currently serving as a Career Development and Mentoring Manager within the Organization for Human Brain Mapping Student and Postdoc Special Interest Group.

Open MRI analysis pipelines with GiraffeTools

Abstract: manuscript_v6final_revsV4_FINAL_4realthistime.pdf. Do the titles of your manuscript ever look like this? And maybe worse, do your scripts and code ever look like this? A major challenge to reproducible work is the handling of versions of scripts, code, and documents. In this talk, we will discuss how to deal with code that is continuously updated; how version control and GitHub can be used to make projects more manageable. We will further explore how this can subsequently be used to collaborate more easily with others, get connected to others, and to work more reproducibly. In addition, we will connect this to ways of better communicating neuroimaging data, results, and pipelines with a variety of open source software.

Bio: Tim is a PostDoc at the Donders Institute for Brain, Cognition, and Behaviour. He has developed tools for making laminar fMRI possible and is particularly interested in workflow development, open science, open source software, and good coding practices. Among others, he has created the Porcupine visual programming application and an Augmented Reality brain visualisation website, ARmadillo. He is also appointed as the chair of the Open Science Room at the next OHBM conference in Rome, 2019.

BIDS in practice for Ultra-High field structural and functional MRI: how to extend an existing standard

Abstract: Most researchers would like to apply open data standards, because such standards make it easier to share data, as well as using tools developed by other groups. The BIDS specification offers a well-defined template for organizing many standard fMRI datasets, but it currently has no clear way of organizing more complex structural, multimodal MRI data common in Ultra-High field MRI, like MP2RAGE images. It is also ambivalent about more complex functional MRI sequences, like multi-shot 3D EPIs. To use BIDS in our work on laminar-resolved 7 Tesla fMRI, we needed to extend BIDS so it could represent multimodal images and its relevant scanning parameters. In this talk, we will lay out what kind of challenges we faced in implementing this extension, such as dealing with unfortunate choices that were made earlier in the specification process, conflicting feedback from the community, and terms that mean slightly different things in different fields. We will also lay out the current draft of the BIDS extension proposal and how you can contribute.

Bio: Gilles de Hollander is a postdoctoral research fellow at the Vrije Universiteit, Amsterdam. He obtained a master’s degree in Artificial Intelligence at the University of Amsterdam in 2012 and a PhD in cognitive neuroscience in 2018, also at the University of Amsterdam. His thesis work was on structural and functional imaging of the subthalamic nucleus at 7 Tesla. As a postdoc, he works in close collaboration with researchers at the Chinese Academy of Sciences and studies binocular rivalry using cortical-depth dependent fMRI at 7 Tesla. His main research interests are UHF-MRI, neuromodulation and its link to perception, as well as open-source software development.

MR image reconstruction using open source software

Abstract: Research into accelerated MRI took a high flight in the past few years. Initially through Compressed Sensing and now via Deep Learning, MRI imaging can be accelerated beyond what was traditionally feasible, thereby reducing scanning time. In this talk we will give background and examples of these approaches. We will then go into software solutions for offline reconstruction and migrating these solutions to the scanner.

Bio: Matthan’s research is on rapid and comprehensive MRI in the human brain. He is currently working on quantitative MRI at high field, diffusion MRI and deep learning for image reconstruction, and is applying these methods in brain disease studies using machine learning.

Open Science in practice

Abstract: Open Science is the ultimate embodiment of the scientific spirit. Jointly, and in a sharing, transparent and critical environment, we generate knowledge for the good of science and society. The reality is often slightly harsher and there are many hurdles to be taken to realise something we all recognise as Open Science. Realising Open Science is also an agile process, a road to be travelled in many small steps, and sometimes stepping back or taking a little D-tour. There are existing infrastructures like DANS, 4TU DataCenter or the EUDAT services, that can help you making your data open and FAIR (Findable - Accessible - Interoperable - Reusable). There are also scientific data initiatives that were /are FAIR and open “avant la lettre”, and act as a guide for further Open Science development in other disciplines. Overall data repositories fulfil a crucial role in the Open Science landscape: They curate and preserve research data, aiding the reproducibility of completed research and serving as a source of data and information for entirely new research. In our presentation, we will share our vision and experiences on Open Science from a repository perspective. We will give an overview of the available infrastructure in the Netherlands and beyond that can aid researchers to make their data FAIR and open. In this way we would like to support the thoughts and developments around Open Data in the magnetic resonance imaging research community.

Bio: Ricarda Braukmann is working as program leader for the Social Sciences at Data Archiving and Networked Services (DANS). She is part of the policy and communication department and involved in various projects promoting Open Science and sustainable Data Management. Ricarda has, for instance, been involved in the creation of an online Research Data Management training module for early career scientists (see online). Ricarda has a multi-disciplinary background in Psychology (BSc) and Cognitive Neuroscience (MSc), and she received her PhD in Developmental Cognitive Neuroscience from the Radboud University Nijmegen in the beginning of 2018.

Cees Hof is working as a cross-disciplinary project acquisition manager at Data Archiving and Networked Services (DANS), an institute of the Royal Netherlands Academy of Arts and Sciences (KNAW). Cees is involved in projects around the European Open Science Cloud (EOSC), software sustainability and Research Data Management (RDM) training and acts as a DANS liaison for the life sciences. Before working for DANS, he was the coordinator of the European Network for Biodiversity Information (ENBI) and thereafter, for more than 10 years, coordinator of the Dutch branch of the Global Biodiversity Information Facility (GBIF). His experience covers all phases of the data life cycle, with a special interest in metadata, geodata, and the pros and cons of data re-use in science and society. Cees has a background in aquatic ecology and ecotoxicology (MSc) and moved into animal systematics, taxonomy, palaeontology and geochemistry for his PhD research at the University of Amsterdam and postdoctoral research in the UK (Bristol).