Rohan Banerjee

I'm a Research Engineer at Montreal Heart Institute, developing multimodal foundation models that align biosignals (primarily ECG) with LLMs for clinically faithful interpretation, structured reporting, and QA. I am also building LLM-as-a-judge evaluation frameworks to measure factuality, consistency, and safety in real cardiology workflows.

I was a Research Master's grad at Mila - Quebec AI Institute and Polytechnique Montreal co-advised by Prof. Julien Cohen-Adad and Prof. Benjamin De Leener where I led the development of EPISeg, a deep learning model for automated segmentation of the spinal cord on gradient-echo EPI fMRI, and co-curated the largest open-access multi-center spinal cord fMRI segmentation dataset on OpenNeuro.

▶ Research

Presentations and Publications

2025

Towards Clinically Faithful ECG Reports via Quantization-Based Tokenization

R. Banerjee, J. Delfrate, R. Avram

Brain and Body Foundation Models Workshop @ NeurIPS 2025

EPISeg: Automated segmentation of the spinal cord on echo planar images using open-access multi-center data

R. Banerjee, M. Kaptan, ....., K. Weber II, B. De Leener, J. Cohen-Adad

Imaging Neuroscience; ISMRM 2025 [Oral Power Pitch, Magna cum Laude Award (Top 10%)]; Dataset; Code & Weights

2023

Automatic spinal cord segmentation based on fMRI EPI data

R. Banerjee, M. Kaptan, K. Weber II, B. De Leener, N. Kinany, F. Eippert, J. Finsterbusch, D. Van De Ville, J. Cohen-Adad

QBIN Scientific Day 2023

Multimodal Pediatric Spinal Cord Template

N. Blostein*, R. Banerjee*, S. Bédard, S. Shahrampour, B. De Leener, F. B. Mohamed, M. Laura Krisa, J. Cohen-Adad

QBIN Scientific Day 2023; Code

2021

Unsupervised annotation of differences between 3D genomic datasets using deep neural networks

E. Shemsu*, R. Banerjee*, A. Raheem

BlackAIR Summer Research Grant Program, Stanford University

2020

Improving the performance of deep convolutional neural networks (CNN) in embryology using synthetic machine-generated images

M. Kanakasabapathy, C. Bormann, P. Thirumalaraju, R. Banerjee, H. Shafiee

Human Reproduction Vol. 35, pp. I209-I209, 2020
* also presented as a poster @ Discover Brigham, HMS (2020)

▶ Past Experience

I was a Research Assistant at Shafiee Lab @ Harvard Medical School, where I developed AI based systems for Assisted Reproductive Technologies under the supervision of Dr. Hadi Shafiee.

I earned my Bachelors in Computer Science from SRM University in India in 2020.