 |
NeurosalienceAuthor: OHBM
The Neurosalience podcast is supported by the Organization for Human Brain Mapping (OHBM). Dr. Peter Bandettini interviews neuroscientists who measure, map, and model brain function and structure and delves into latest advancements, challenges, controversies, and controversies. He engages young and old and strives to add insight and perspective wherever the conversation goes. Language: en Genres: Life Sciences, Science Contact email: Get it Feed URL: Get it iTunes ID: Get it |
Listen Now...
Neurosalience #S6E10 with Satrajit Ghosh - How Better Tools Can Transform Brain Science
Episode 10
Sunday, 29 March, 2026
“Shortening scientific loops accelerates discovery”Dr. Satrajit Ghosh is a senior research scientist at the McGovern Institute for Brain Research at MIT and an assistant professor at Harvard Medical School. He has helped advance neuroinformatics, open science, and reproducible neuroimaging through both his research and the development of widely used community tools. His work spans machine learning for neuroimaging, the neural mechanisms of speech, and the use of speech features to inform psychiatric diagnosis and treatment. He earned his bachelor’s degree with honors in computer science from the National University of Singapore and his PhD in cognitive and neural systems from Boston University. He has contributed to influential projects including Nipype, fMRIPrep, and NeuroVault. More recently, he has focused on how shared scientific infrastructure can connect domains, modalities, and scales across neuroscience and help address the field’s growing fragmentation.In this episode, Peter and Satrajit discuss the origins of tools like Nipype and the broader push for reproducible neuroimaging, showing how practical research challenges can inspire infrastructure that benefits the entire field. They also explore functional gradients in the brain and cerebellum, the promise of speech as a scalable biomarker for mental health, and the cautious role AI may play in diagnosis and scientific discovery. A major theme in their conversation is the fragmentation in neuroscience, with knowledge often siloed across methods, scales, and communities. Ghosh argues for a more intelligent scientific infrastructure that connects data, tools, theory, and expertise. He closes with advice to young scientists: experiment often, make mistakes, and learn by discovering where systems fail.We hope you enjoy this episode!Chapters00:00 Introduction to Satra Ghosh and His Work06:46 The Intersection of Control Theory and Speech11:18 Satra’s Academic Journey into Neuroscience20:58 Neuroinformatics and Tool Development34:42 Individual Differences in Brain Structure39:21 Developing tools to augment Experimental Design44:25 Building an Intelligent Infrastructure for Neuroscience58:45 The Role of Theory in Neuroscience01:00:26 Access to Scientific Research Expediting Progress01:06:40 Experience Inherent to Learning 01:09:33 Mapping the Brain’s Functional Gradient01:16:31 AI and Speech Analysis in Mental Health01:29:31 Advice, Fail More, Learn MoreWorks mentioned:34:59 - Marek, S. et al. (2022). Reproducible brain-wide association studies require thousands of individuals. https://doi.org/10.1038/s41586-022-04492-943:44 - Ghosh, Satrajit (2025). An Intelligent Infrastructure as a Foundation for Modern Science.https://doi.org/10.48550/arXiv.2508.1005101:09:33 - Margulies, Daniel S., et al. (2016). Situating the default-mode network along a principal gradient of macroscale cortical organization. https://doi.org/10.1073/pnas.160828211301:10:13 - Xavier Guell, Jeremy D Schmahmann, John DE Gabrieli, Satrajit S Ghosh (2018). Functional gradients of the cerebellum. https://doi.org/10.7554/eLife.36652Tools and resources mentioned:Nipype : an open-source Python framework for building reproducible neuroimaging workflows.https://nipype.readthedocs.io/en/latest/index.htmlfMRIPrep : a robust, analysis-agnostic preprocessing pipeline for functional MRI. https://fmriprep.org/en/stable/OpenScope : an open-science effort for large-scale neuroscience data sharing and analysis.https://www.allenneuraldynamics.org/projects/openscopeDANDI : a platform for publishing, sharing, and processing neurophysiology data. https://about.dandiarchive.org/NeuroVault : A public repository of unthresholded statistical maps, parcellations, and atlases of the brain.https://neurovault.org/Episode producers:Ömer Faruk Gülban, Karthik Sama
