Hongyi Duanmu 端木弘毅

I am a 5th-year PhD student at the Department of Computer Science, Stony Brook University, under the supervision of Prof. Fusheng Wang. Before that, I received my B.E. degree from the University of Science and technology of China under the supervision of Prof. Yan Song.

My research interests focused on medical image analysis, especially with deep learning techniques.

Email  /  CV  /  Google Scholar  /  Github

profile photo
blind-date Spatial Attention-based Deep Learning System for Breast Cancer Pathological Complete Response Prediction with Serial Histopathology Images in Multiple Stains
Hongyi Duanmu, Shristi Bhattarai, Hongxiao Li, Chia Cheng Cheng, Fusheng Wang, George Teodoro, Emiel A.M. Janssen, Keerthi Gogineni, Preeti Subhedar, Ritu Aneja, Jun Kong
International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI), 2021

A deep learning-based system was presented to predict PCR to neoadjuvant chemotherapy for triple negative breast cancer patients with multi-stained histopathology images. Spatial attention mechanism was deployed to catpure cell type, shape, and location information for enhanced PCR prediction

blind-date Foveal blur-boosted segmentation of nuclei in histopathology images with shape prior knowledge and probability map constraints
Hongyi Duanmu, Fusheng Wang, George Teodoro, Jun Kong
Bioinformatics, 2021

To contour cell in pathology images, we proposed a CNN with foveal blur enriching datasets with multiple local nuclei regions of interest. A human-knowledge boosted deep learning system was proposed by including new loss function terms capturing shape prior knowledge and imposing smoothness constraints on the predicted probability maps.

blind-date Prediction of Pathological Complete Response to Neoadjuvant Chemotherapy in Breast Cancer Using Deep Learning with Integrative Imaging, Molecular and Demographic Data
Hongyi Duanmu, Pauline Boning Huang, Srinidhi Brahmavar, Stephanie Lin, Thomas Ren, Jun Kong, Fusheng Wang, Tim Q. Duong
International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI), 2020

We propsoed a novel approach integrating 3D MRI imaging data, molecular data and demographic data using convolutional neural network to predict the likelihood of pathological complete response to neoadjuvant chemotherapy in breast cancer.

blind-date Automatic Brain Organ Segmentation with 3D Fully Convolutional Neural Network for Radiation Therapy Treatment Planning
Hongyi Duanmu, Jinkoo Kim, Praitayini Kanakaraj, Andrew Wang, John Joshua, Jun Kong, Fusheng Wang
International Symposium on Biomedical Imaging (ISBI), 2020

3D organ contouring is an essential step in radiation therapy treatment planning for organ dose estimation and plans optimizition to reduce organs-at-risk doses. BrainSegNet, a 3D fully convolutional neural network (FCNN) based approach, was presented for automatic segmentation of brain organs.

blind-date Detection of Suicidality Among Opioid Users on Reddit: Machine Learning–Based Approach
Hannah Yao, Sina Rashidian, Xinyu Dong, Hongyi Duanmu, Richard N Rosenthal, Fusheng Wang
Journal of Medical Internet Research, 2020

This study aimed to extract posts of suicidality among opioid users on Reddit using machine learning methods. The analysis will help us to better understand the rationale of these users, providing new insights into individuals who are part of the opioid epidemic.

blind-date Convolutional Neural Network Detection of Axillary Lymph Node Metastasis Using Standard Clinical Breast MRI
Thomas Ren, Renee Cattell, Hongyi Duanmu, Pauline Huang, Haifang Li, Rami Vanguri, Michael Z. Liu, Sachin Jambawalikar, Richard Ha, Fusheng Wang, Jules Cohen, Clifford Bernstein, Lev Bangiyev, Timothy Q. Duong
Clinical Breast Cancer, 2020

A convolutional neural network was constructed to noninvasively predict axillary lymph node metastasis from standard clinical breast MRI at the pre-neoadjuvant chemotherapy stage.


Reviewer for MICCAI, ISBI, JBHI, JBI

Design and code stolen from Jon Barron