Hanyi Fang

Dr. Hanyi Fang

orcid

Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
12 July 2021 in Applied Sciences
Verified
Journal Article
Open access
PDF file
DJDi Jin
Di Jin
EPEileen Pan
Eileen Pan
NONassim Oufattole
Nassim Oufattole
WWWei-Hung Weng
Wei-Hung Weng
HFHanyi Fang
Hanyi Fang
PSPeter Szolovits
Peter Szolovits
Open domain question answering (OpenQA) tasks have been recently attracting more and more attention from the natural language processing (NLP) community. In this work, we present the first free-form multiple-choice OpenQA dataset for solving medical problems, MedQA, collected from the professional medical board exams. It covers three languages: English, simplified Chinese, and traditional Chinese, and contains 12,723, 34,251, and 14,123 questions for the three languages, respectively. We implement both rule-based and popular neural methods by sequentially combining a document retriever and a machine comprehension model. Through experiments, we find that even the current best method can only achieve 36.7%, 42.0%, and 70.1% of test accuracy on the English, traditional Chinese, and simplified Chinese questions, respectively. We expect MedQA to present great challenges to existing OpenQA systems and hope that it can serve as a platform to promote much stronger OpenQA models from the NLP community in the future.

21 May 2021 in Preprints.org
Verified
Preprint
Open access
PDF file
DJDi Jin
Di Jin
EPEileen Pan
Eileen Pan
NONassim Oufattole
Nassim Oufattole
WWWei-Hung Weng
Wei-Hung Weng
HFHanyi Fang
Hanyi Fang
PSPeter Szolovits
Peter Szolovits
Open domain question answering (OpenQA) tasks have been recently attracting more and more attention from the natural language processing (NLP) community. In this work, we present the first free-form multiple-choice OpenQA dataset for solving medical problems, MedQA, collected from the professional medical board exams. It covers three languages: English, simplified Chinese, and traditional Chinese, and contains 12,723, 34,251, and 14,123 questions for the three languages, respectively. We implement both rule-based and popular neural methods by sequentially combining a document retriever and a machine comprehension model. Through experiments, we find that even the current best method can only achieve 36.7%, 42.0%, and 70.1% of test accuracy on the English, traditional Chinese, and simplified Chinese questions, respectively. We expect MedQA to present great challenges to existing OpenQA systems and hope that it can serve as a platform to promote much stronger OpenQA models from the NLP community in the future.

7 March 2025 in Journal of Nanobiotechnology
Verified
Journal Article
Open access
PDF file
XWXingyi Wang
Xingyi Wang
HFHanyi Fang
Hanyi Fang
WHWenzhu Hu
Wenzhu Hu
YFYuan Feng
Yuan Feng
ZZZhangyongxue Zhou
Zhangyongxue Zhou
XLXIAOLI LAN
XIAOLI LAN
Limited treatment response and inadequate monitoring methods stand firmly before successful immunotherapy. Recruiting and activating immune cells in the hypoxic tumor microenvironment is the key to reversing immune suppression and improving immunotherapy efficacy. In this study, biomimetic oxygen-delivering nanoparticles (CmPF) are engineered for homologous targeting and hypoxia alleviation within the tumor environment. CmPF targets the tumor microenvironment and delivers oxygen to reduce hypoxia, thereby promoting immune cell activity at the tumor site. In addition, granzyme B-targeted positron emission tomography (PET) imaging is employed to monitor immune cell activity changes in response to immunotherapy efficacy in vivo. The combination of CmPF with carboplatin and PD-1 inhibitors significantly suppresses tumor growth by 2.4-fold, exhibiting the potential of CmPF to enhance the efficacy of immunotherapy. Immunohistochemistry further confirms increased expression of key immune markers, highlighting the reprogramming of the tumor microenvironment. This study demonstrates that hypoxia alleviation enhances tumor immunotherapy efficacy and introduces a non-invasive PET imaging method for dynamic, real-time assessment of therapeutic response.

12 January 2022 in Molecular Imaging and Biology
Verified
Journal Article
Closed access
PDF file
HFHanyi Fang
Hanyi Fang
SRSamantha Rossano
Samantha Rossano
XWXingxing Wang
Xingxing Wang
NNNabeel Nabulsi
Nabeel Nabulsi
BKBrian Kelley
Brian Kelley
YHYiyun Huang
Yiyun Huang

1 April 2026 in Radiology
Verified
Journal Article
Closed access
CTCheng Tang
Cheng Tang
XSXun Sun
Xun Sun
ATAnqi Tang
Anqi Tang
WRWeiwei Ruan
Weiwei Ruan
FLFang Liu
Fang Liu
FIFor The Alzheimer’S Disease Neuroimaging Initiative
For The Alzheimer’S Disease Neuroimaging Initiative
An interpretable, Alzheimer disease–specific pathophysiologic-constrained adversarial decomposition framework yielded voxel-level pathologic maps and an artificial intelligence biomarker with stronger cross-sectional and temporal associations with cognition and hippocampal volume than standard PET biomarkers.

30 December 2025 in Clinical Nuclear Medicine
Verified
Journal Article
Open access
XWXingyi Wang
Xingyi Wang
YLYuhu Lv
Yuhu Lv
CQChunxia Qin
Chunxia Qin
XLXIAOLI LAN
XIAOLI LAN
HFHanyi Fang
Hanyi Fang
We present a rare case of an isolated splenic myofibroma in a child evaluated with 18 F-FDG PET/CT. An 8-year-old boy underwent PET/CT for a left cervical mass, which incidentally revealed a splenic nodule (1.3×1.0 cm) with markedly increased 18 F-FDG uptake (SUV max =8.1). MRI showed an enhancing, low T2-weighted signal nodule without diffusion restriction. Histopathology after splenectomy confirmed myofibroma. Isolated myofibromas are rare benign spindle cell tumors in children and can demonstrate intense 18 F-FDG uptake, mimicking malignancy. Awareness of this potential false-positive pitfall is essential in pediatric PET interpretation.