Announcing the Quality Conversations webinar series [on niQC, open science etc]

Update - events in reverse chronological order: Aug 12, 11am EST : Prof. Tonya White on Automatic Quality Assessment of Structural Images in Pediatric Neuroimaging data, presenting this paper White, T., et al. (2018). Automated quality assessment of structural magnetic resonance images in children: Comparison with visual inspection and surface-based reconstruction. Human Brain Mapping, 39(3), 1218–1231. https://doi.org/10.1002/hbm.23911July …

Continue reading Announcing the Quality Conversations webinar series [on niQC, open science etc]

Video recordings of talks and tutorials from Pitt BrainHack DeBurghing 2020

The video recordings of keynotes, talks and tutorials from DeBurghing 2020 are now posted on youtube via the Open MINDS Lab channel in this playlist below (or you can use this direct URL). Apologies for the delay in posting them, and I thank you for your patience. Many thanks to all the speakers and participants …

Continue reading Video recordings of talks and tutorials from Pitt BrainHack DeBurghing 2020

The inaugural Pitt Brain Hack: DeBurghing 2020 [hackathon]

I am super excited to announce the inaugural Pitt BrainHack The @DeBurghing 2020 on Dec 9-11 with a ton of exciting talks and projects on machine learning, handling confounds in predictive analysis, multi-site data harmonization, data science, high-performance computing (HPC) and the like. It's completely free and open to everyone in the world. Join us …

Continue reading The inaugural Pitt Brain Hack: DeBurghing 2020 [hackathon]

Featured

Announcing the Open Minds lab @ Pitt

I am super excited to announce I am joining the University of Pittsburgh as an Assistant Professor in the Department of Radiology. At the new Open minds in Machine learning, INformatics and Data Science (MINDS) lab (follow us at @OpenMindsLab), we aim to solve high-impact biomedical challenges leveraging our expertise in medical image analysis, signal …

Continue reading Announcing the Open Minds lab @ Pitt

My goals to further Diversity and Inclusion in Science

Being a visible minority and often underrepresented, I know the importance of improving diversity and inclusivity, in all the important dimensions including but not limited to gender, sexual orientation, ethnicity, race, culture, geography and school of thought. My lived experience in this domain taught me a lot of things, and provided me an important perspective …

Continue reading My goals to further Diversity and Inclusion in Science

Evaluating Visual QC protocols for Freesurfer Parcellations: error detection rate, inter-rater reliability and site differences

This page will host the study and results from systematically evaluating the visual QC protocols for Freesurfer parcellations. A preprint is available here. A poster summarizing the results presented at OHBM 2020, titled "Site-differences and inter-rater reliability for visual quality control (QC) for Freesurfer Parcellations" is shown below. The VisualQC tool is available at here. You …

Continue reading Evaluating Visual QC protocols for Freesurfer Parcellations: error detection rate, inter-rater reliability and site differences

Unambiguous terminology for data-splits in nested cross-validation (CV): training, tuning and reporting sets

A particularly frustrating confusion exists when referring to the different dataset splits when performing nested CV: whether the split in [the really] outer CV loop is to be called test set or validation set, and whether the "test set" in the inner-CV loop is to be called validation set or test set. There are different …

Continue reading Unambiguous terminology for data-splits in nested cross-validation (CV): training, tuning and reporting sets

Research data management for medical data with pyradigm

Here are the slides from my talk the US Research Software Engineering (RSE) association workshop, titled "Research data management for medical data with pyradigm". pyradigm is a Python-based data structure for biomedical data to manage multiple tables (of diverse data types) linked via common patient info (or any other hashable ID). Allowing continuous validation, this …

Continue reading Research data management for medical data with pyradigm

Conquering confounds and covariates in machine learning

News: here is the video pitch I've made for the OHBM hackathon 2020. The associated slides are here. As I tried to study impact of different deconfounding methods, as well as offer covariate regression ability in neuropredict, I realized the tools and methods I need to implement would be useful to broader machine learning and …

Continue reading Conquering confounds and covariates in machine learning