1248
Educational Course - Half Day (4 hours)
The use of naturalistic conditions (e.g., movies, story listening) is expanding rapidly in the neuroimaging community, and many groups with no previous experience using these complex stimuli are now incorporating naturalistic conditions into experiments. There are now also multiple open-science databases available that include naturalistic data (e.g., HCP 7T release, Healthy Brain Network, Narratives Dataset), so many researchers are conducting analyses on naturalistic data with increasing frequency. Though the scanning sequence and data format looks the same as resting state or conventional task data, using these dynamic, rich stimuli involves challenges and considerations that are unique. This half-day course will provide a well-rounded overview of active and emerging issues across a broad range of topics, from stimuli selection to data processing and analytics, to sharing of both stimuli and data. It also provides a bidirectional opportunity for the community to help shape the best practice recommendations of our working group, thereby furthering the utility of future recommendations and the practice of rigorous neuroscience within this rich, complex space.
The learning outcomes are that each participant will be able to articulate a conceptual framework for working with naturalistic stimuli, will know emerging best practice recommendations across six topics, will have a list of leading-edge open-science resources relevant to the work, and will have contributed at a key developmental phase to the best practices workgroup.
1. Describe a conceptual framework for working with time-locked naturalistic stimuli, with a focus on factors and trade-offs that are unique to naturalistic conditions. List key papers that are foundational to the evolution of the subfield.
2. Hands-on use of a representative data-sharing platform (“Naturalistic Imaging Data Catalogue). Audience members will independently interact with the platform and navigate through a data sharing tutorial in real-time.
3. Resources and tools: Know where to find up-to-the-minute resources that relate to naturalistic neuroimaging, including stimuli, analytic pipelines, shared code, and where, how and why to share both stimuli and data.
4. Best practice recommendations: Identify two key practice recommendations for each phase of a naturalistic experiment, and provide input to the workgroup about common questions, areas of confusion, or upcoming and future challenges.
The target audience includes novice, experienced and future users of naturalistic stimuli. We anticipate that this will include researchers across the entire lifespan, from both clinical and non-clinical research programs, and researchers doing both human and non-human primate work. Though the focus will be on functional MRI techniques, researchers that use EEG, MEG, and fNIRs, etc. will also benefit from the content and resources.
Presentations
This lightning talk (8 minutes) will provide the community with an overview of the major issues related to preprocessing data that are specific to naturalistic conditions. The current drafts of best practice recommendations will be outlined, and links to relevant resources/open-science tools/major papers will be provided. Finally, outstanding questions or challenges will be raised to prompt discussion and interactions during the subsequent roundtables.
Presenter
Bharat Biswal, New Jersey Institute of Technology Newark, NJ
United States
This 8-minute talk will provide an overview of the major issues related to selecting and using naturalistic stimuli, including how to select stimuli for an experiment, how to describe and report stimuli in papers, how to characterize or annotate the statistics of stimuli themselves, and how to characterize subjective and physiological responses to stimuli. The current drafts of best practice recommendations will be outlined, and links to stimulus-based resources will be highlighted. The focus will be on practical, hands-on approaches to using naturalistic stimuli.
Presenter
Peter Kirk, NIH Bethesda, MD
United States
One of the major advantages of naturalistic neuroscience is that these datasets enable a panoply of analytic approaches, spanning from task-based GLMs to conventional resting state analyses to unique inter subject approaches. Within this broad array of options, even the most straightforward approaches require unique considerations when the data are naturalistic. This brief lecture will provide audience members with a schematic to understand the analytic options (including statistical issues), and will highlight some key best practice recommendations. The talk will end by outlining ongoing active questions and challenges to spur interactive discussion.
Presenter
Saurabh Sonkusare, The University of Newcastle, Australia Newcastle
Australia
Data sharing and open-science practices have a strong lineage within the naturalistic neuroimaging community. This stems in part from the data themselves, including the way they can be used by multiple groups, and the close interdigitation between stimuli and data. This 8-minute presentation aims to enable both new and experienced researchers to start to engage in leading edge data sharing practices and resources. This (and the subsequent roundtable) will cover both conceptual and technical issues related to the sharing of stimuli, data, metadata and code.
Presenter
Sam Nastase, Princeton University Princeton, NJ
United States
This brief presentation will present one specific data sharing platform ("The Naturalistic Data Imaging Catalogue") and review its structure and features for the audience. The idea is to provide a hands-on introduction to one example platform that has been specifically designed to handle both stimuli and data from naturalistic experiments. Strengths and limitations, as well as future features, will be reviewed. In the subsequent roundtable, interested audience members will have the opportunity to work on the catalogue in real time, ask any questions specific to their data needs, and to offer feedback.
Presenter
Susanne Weis, Research Centre Jülich
Research Centre Jülich
Jülich, n/a
Germany
When using complex narrative stories and films as scientific stimuli, a myriad of important considerations must be taken into account to conduct rigorous and responsible science. These include social and cultural factors, consideration of equity, diversity and inclusion principles in the stimuli themselves, developmental and lifestyle factors, and illness- or population-specific issues. This presentation will provide a framework for thinking about these issues, and review emerging best practice recommendations. The following roundtable session will allow participants to ask questions specific to their own work, and to contribute to what needs to be a broad, community-informed, evolving discussion.
Presenter
Signe Bray, University of Calgary Calgary, Alberta
Canada