Mahmudul Hasan

I am a PhD candidate at Stony Brook University, working with Prof. Chao Chen and Prof. Joel Saltz. My research focuses on developing interpretable machine learning models and advanced spatial statistics tools for medical imaging, aiming to enhance the trustworthiness and accuracy of AI in healthcare. I leverage probabilistic models, such as contrastive Variational Autoencoders (cVAEs), to disentangle meaningful factors in medical images, improving explainability and clinical reliability. Additionally, I have created an interactive spatial analysis toolkit that enables pathologists to test spatial hypotheses and uncover complex relationships within tissue samples, leading to novel insights, such as identifying spatial phenotypes with significant survival differences across cancer subtypes. My work integrates foundational models, multimodal analysis, and collaborative approaches to address critical challenges at the intersection of AI and medicine.

news

August 2024	Our work on the New Spatial Phenotypes from Imaging Uncover Survival Differences for Breast Cancer Patients was accepted at ACM BCB 2024 for ORAL presentation.

July 2024	Our work on the Semi-supervised contrastive VAE for disentanglement of digital pathology images was accepted at MICCAI 2024.

publications

Semi-Supervised Contrastive VAE for Disentanglement of Digital Pathology Images

Mahmudul Hasan, Xiaoling Hu, Shahira Abousamra, Prateek Prasanna, Joel Saltz, Chao Chen

In MICCAI, 2024

TL;DR: Can we disentangle meaningful factors from pathology images?

Abs arXiv Bib Code

Despite the strong prediction power of deep learning models, their interpretability remains an important concern. Disentanglement models increase interpretability by decomposing the latent space into interpretable subspaces. In this paper, we propose the first disentanglement method for pathology images. We focus on the task of detecting tumor-infiltrating lymphocytes (TIL). We propose different ideas including cascading disentanglement, novel architecture, and reconstruction branches. We achieve superior performance on complex pathology images, thus improving the interpretability and even generalization power of TIL detection deep learning models.
@inproceedings{hasan2024semi, title = {{Semi-supervised contrastive VAE for disentanglement of digital pathology images}}, author = {Hasan, Mahmudul and Hu, Xiaoling and Abousamra, Shahira and Prasanna, Prateek and Saltz, Joel and Chen, Chao}, booktitle = {International Conference on Medical Image Computing and Computer-Assisted Intervention}, year = {2024}, }