Big Data Neuroscience Workshop 2019: Organized by the Advanced Computational Neuroscience Network (ACNN)

Overview

About

Organizers: A team of transdisciplinary investigators from the Advanced Computational Neuroscience Network (ACNN), including:

• University of Michigan: Ivo Dinov, Rich Gonzales, John Marcotte
• Indiana University: Franco Pestilli, Olaf Sporns, Andrew Saykin
• OSU: Dhabaleswar Panda, Xiaoyi Lu, Hari Subramoni
• CWRU: Satya Sahoo
• Northwestern University: Lei Wang

Objectives: The ACNN 2019 workshop will continue our work on the development of common practices and standardization to make it easier for neuroscience researchers to annotate and process data; to share data, tools and protocols, and to work with distributed high-performance computing environments. The workshop will bring together members of the Midwest, national, and the global neuroscience research community to promote data reuse, aggregation, result validation and new discoveries in neuroscience.

Program

Big Data Neurosciences Workshop 2019

Organized by the Advanced Computational Neuroscience Network and supported by the National Science Foundation

Keynote Speakers

Russ Poldrack

Frontiers in Brain Imaging

Title:

Building around standards: The role of BIDS in neuroimaging data sharing

Description:

There is increased interest in data sharing, but shared data on their own are not useful unless they follow standard structures that allow them to be easily interpreted and reused. I will discuss the development of the Brain Imaging Data Structure (BIDS), a community effort to describe neuroimaging datasets. I will show how the availability of this standard has enhanced the ability to automatically share and process neuroimaging datasets, focusing on the OpenNeuro and BIDS Apps projects.

Tristan Glatard

Title:

Big Data Infrastructures for Neuroinformatics: Performance, Reproducibility, Interoperability

Description:

Big Data engines such as Apache Spark, Hadoop or Dask, are now widely adopted in various industry domains, but they remain seldom used in neurosciences while large data volumes are commonly processed. The first part of this talk will explain the reasons for this situation. More precisely, it will review (1) what are the main properties of Big Data engines and how they differ from traditional neuroimaging engines such as Nipype, and (2) in which situations Big Data engines provide a performance advantage to neuroimaging applications. In a second part, the talk will highlight the consequences of using Big Data infrastructures on the reproducibility of neuroimaging results, taking examples from structural and functional MRI analysis. In particular, it will show the effect of minor data and infrastructural perturbations on analysis results. Finally, the presentation will conclude by demonstrating a few technological solutions to allow Big Data platforms to interoperate, in particular the Boutiques framework for pipeline interoperability.

Jessica Turner

Title:

Neuroinformatics and imaging genetics: ENIGMA & COINSTAC

Description:

Large-scale neuroimaging studies and analysis can be implemented in a number of ways, from a planned multi-site effort in data collection to opportunistic analyses of legacy data. The Enhancing Neuroimaging Genetics through Meta Analysis (ENIGMA) consortium is currently in its 10th year of facilitating large-scale data sharing and analysis techniques to produce analyses of thousands of participants' data, through a combination of surprisingly simple techniques. I will provide an overview of how ENIGMA functions and its methods; as an open consortium, it is relatively easy to get involved. ENIGMA runs into limits in what data shared for aggregated analyses, particularly for genetics; the Collaborative Informatics and Neuroimaging Suite Toolkit for Anonymous Computation (COINSTAC) has been developed as a platform to allow decentralized analyses in a context of privacy protection for sensitive data. I will present some of the examples of COINSTAC's capabilities and how it interacts with ENIGMA and other ad hoc collaborations.

Jack Van Horn

Department of Psychology and School of Data Science, University of Virginia

Title:

Data Science: An Introduction to the Computational Tools and Techniques for Large-scale Biomedical Research

Description:

It has been noted that data science and large-scale analyses have ushered in a wholly new research paradigm. With the onslaught of rich biomedical data types related to the brain form, function, and connectivity, as well as to aspect such as neurological disease and drug abuse, these methods form the basis for several lines of modern neuroscientific examination. In this presentation, I will review the motivations behind data science, several computational tools, and feature examples of where and how it has utility for the ‘big data’ examination of the human brain. Important in this is the education and training of the next generation of biomedical data scientists.

Lightning Talks

Posters

Poster dimensions – 48x72 (Horizontal Layout)

Posters will be on display throughout the conference with a reception and poster session at the end of the first day of the conference, Thursday, September 19, 2019

List of Accepted Posters

• Poster 1: Spatial-temporal perturbations in ADAR editing as prognostic and diagnostic biomarkers in neurological disorders
  by Noel-Marie Plonski, Caroline Nitirahardjo, Ain Shajihan and Helen Piontkivska
• Poster 2: Phase coupling and amplitude coupling are two distinct but spatially and temporally associated mechanisms of connectivity; An intracranial EEG study
  by Parham Mostame, Abbas Babajani-Feremi and Sepideh Sadaghiani
• Poster 3: Predicting ICH Patient Outcome from Brain CT Scans Using an Ensemble Deep Learning Framework
  by Balaji Iyer, Smruti Deoghare, Samuel Hacker, Vivek Khandwala, David Wang, Daniel Woo, Achala Vagal and Surya Prasath
• Poster 4: Mosaicking and Blending in Large-scale Neuroimaging for Robust Dendrite Detection
  by V. B. Surya Prasath and Bruce Aronow
• Poster 5: Spatial Updating and Domain Expertise: the case of dancers
  by Maria Photiou and Marios Avraamides
• Poster 6: Making network neuroscience more findable, accessible, interoperable, and reusable via brainlife
  by Brent McPherson, Daniel Bullock, Brad Caron, Lindsey Kitchell, Soichi Hayashi, Josh Faskowitz, Olaf Sporns, Richard Betzel, Paolo Avesani and Franco Pestilli
• Poster 7: Data Quality And Safety Of Simultaneous EEG-fMRI Using Multi-band fMRI Sequences
  by Maximillian Egan, Ryan Larsen, Brad Sutton and Sepideh Sadaghiani
• Poster 8: Characterizing and Understanding Time-varying Functional Connectivity States via Network Science and Deep Neural Networks
  by Marlena Duda, Danai Koutra and Chandra Sripada
• Poster 9: Longitudinal effects of dimensional exposure to violence exposure and social deprivation on adolescent white matter tracts
  by Leigh Goetschius, Tyler Hein, Colter Mitchell, Nestor Lopez-Duran, Vonnie McLoyd, Sara McLanahan, Jeanne Brooks-Gunn, Luke Hyde and Christopher Monk
• Poster 10: Smoking Behavior and Cumulative Genetic Risk for Smoking on Cognitive Status among Older Americans
  by Mingzhou Fu, Harita Vadari, Erin Ware and Kelly Bakulski
• Poster 11: Decentralized analysis of brain imaging data: A status update on COINSTAC
  by Harshvardhan Gazula, Bradley Baker, Eswar Damaraju, Sergey Plis, Sandeep Panta, Rogers Silva and Vince Calhoun
• Poster 12: Building Generalizable cognitive models using ECoG data: An improvised technique for group-level analysis
  by Karthik Ganeson, Adriene Beltz and David Brang

Travel Scholarship

Travel Scholarship Awardees:

Sharable Resources

A web form is available to submit items for inclusion in the sharable resources. Examples (not an exclusive list) of appropriate resources that may be suggested includes:

• Highly scalable APIs
• Relevant publications
• Cloud-services
• Computational Resources
• Algorithms, methods, techniques
• Education and Training Opportunities