Arjun Desai

I am a PhD student at Stanford, where I work at the intersection of signal processing, machine learning, and data systems.

I am advised by Akshay Chaudhari and Chris Ré and am affiliated with the Stanford AI Lab, Center for Artificial Intelligence in Medicine and Imaging (AIMI), and Center for Research on Foundation Models (CRFM).

News

08/2023: Delighted that VORTEX-SS received Best Presentation at the NDSEG Fellows conference.
01/2023: Very grateful to be awarded the inaugaral SPORR Research Rigor and Reproducibility Award.
12/2022: I gave a talk at MRITogether on simplifying ML data, development, and deployment interfaces in medical imaging.
10/2022: Excited to talk about democratizing data ecosystems for machine learning at the ISMRM Reproducibility Research meeting
9/2022: I presented our ongoing work on at MedAI on using physics-based priors to robustness and data-efficiency in image reconstruction.
7/2022: Our work VORTEX received best paper at MIDL

Research

I am broadly interested in how we can use machine learning robustly, efficiently and at scale in practice. My research focuses on the intersection of inverse problems in signal processing and machine learning. I also work on how we can build scalable deployment and validation systems for challenging applications in heathcare and the sciences.

For a full list of publications, please see Google Scholar.

VORTEX: Physics-Driven Data Augmentations for Consistency Training for Robust Accelerated MRI Reconstruction
Arjun Desai, Beliz Gunel, Batu Ozturkler, Harris Beg, Shreyas Vasanawala, Brian Hargreaves, Christopher Ré, John Pauly, Akshay Chaudhari
MIDL, 2022 (Oral Presentation, Best Paper)
arXiv | code

SKM-TEA: A Dataset for Accelerated MRI Reconstruction with Dense Image Labels for Quantitative Clinical Evaluation
Arjun Desai, Andrew Schmidt, Elka Rubin, Christopher Sandino, Marianne Black, Valentina Mazzoli, Kathryn Stevens, Robert Boutin, Christopher Ré, Garry Gold, Brian Hargreaves, Akshay Chaudhari
NeurIPS Datasets & Benchmarks, 2021
arXiv | code | dataset | colab

Noise2Recon: Enabling Joint MRI Reconstruction and Denoising with Semi-Supervised and Self-Supervised Learning
Arjun Desai, Batu Ozturkler, Christopher Sandino, Robert Boutin, Marc Willis, Shreyas Vasanawala, Brian Hargreaves, Christopher Ré, John Pauly, Akshay Chaudhari
Accepted (Magnetic Resonance in Medicine), 2023
arXiv | code

The International Workshop on Osteoarthritis Imaging Knee MRI Segmentation Challenge: A Multi-Institute Evaluation and Analysis Framework on a Standardized Dataset
Arjun Desai, Francesco Caliva, Claudia Iriondo, {16 authors}, Akshay Chaudhari.
Radiology: Artifical Intelligence, 2021
arXiv | code

Technical Considerations for Semantic Segmentation in MRI using Convolutional Neural Networks
Arjun Desai, Garry Gold, Brian Hargreaves, Akshay Chaudhari
Preprint, 2019
arXiv | code

Software

Meerkat
Github

Thanks to Jon Barron for this template.