AI for Healthcare
Medical named entity recognition, automatic ICD coding, medical insurance fraud detection, and audit decisions that account for cost risk.
01
About
I am a laboratory engineer at the School of Artificial Intelligence, Wuhan University of Technology. I completed my Ph.D. in Management Science and Engineering at Central South University, advised by Prof. Hongyu Zhang, with a visiting year at the Department of Data Science and Statistics, National University of Singapore.
My research asks a practical question: how can deep learning be made useful — and trustworthy — in settings where a wrong answer carries a real cost? I have built models for Chinese medical named entity recognition, automatic ICD coding, medical insurance fraud detection, and long-horizon forecasting of industrial energy systems.
Much of this work was carried out on real data rather than on benchmarks: a provincial DRG payment reform covering tens of millions of claims and a healthcare fund in the tens of billions of yuan, where the deliverable was more than twenty reports handed to a provincial healthcare security administration.
6journal and conference papers
4Q1 journal articles
2granted invention patents
4national and provincial projects
02
Research
NLP & Text Analytics
Length-diverse entity modeling, cross-lingual dialogue evaluation, and multimodal analysis of user-generated reviews.
Time-Series Forecasting
Flow-enhanced Transformers for long-horizon prediction, fusing time-variant and time-invariant signals.
Data-Driven Policy
DRG payment reform evaluation, healthcare data governance, and evidence that regulators can actually act on.
03
Publications
Journal Articles
01
A Chinese medical named entity recognition method considering length diversity of entities
H. Zhang, L. Lyu, W. Chang, et al. Engineering Applications of Artificial Intelligence, 150: 110649.
02
AI vs. Human: A large-scale analysis of AI-generated fake reviews, human-generated fake reviews and authentic reviews
Y. Zhao, H. Zhang, L. Lyu. Journal of Retailing and Consumer Services, 87: 104400.
03
Can deep learning technology really recognize Mpox? A positive response from the EfficientNet model
X. Zhao, L. Lyu (co-first author), L. Zhang. Frontiers in Microbiology, 16: 1627311.
04
Multimodal user-generated content analysis for digital cultural tourism: An explainable machine learning approach to tourist satisfaction
K. Shen, H. Zhang, L. Lyu (corresponding author). Annals of Tourism Research Empirical Insights.
05
News recognition and classification of emergencies based on a combined deep learning model
Y. Song, L. Lyu, D. Liu. Journal of the China Society for Scientific and Technical Information, 40(2): 145–151. (in Chinese)
Conference
06
An interpretable and cross-lingual method for evaluating second-language dialogues
W. Gao, J. Wu, L. Lyu, et al. Proceedings of NAACL 2025.
Patents
P1
Chinese medical named entity recognition method, apparatus, electronic device and storage medium
H. Zhang, L. Lyu, Y. Zhao.
P2
A medical insurance fraud identification method and system
H. Zhang, L. Lyu, X. Yuan.
04
Projects
01
Coordinated DRG payment reform in the Chang-Zhu-Tan-Heng region, Hunan
Assessed how the reform affected payment efficiency, hospital management quality, and service delivery. Evaluated the accuracy of the underlying claims data, validated the DRG grouping scheme against clinical practice, and built the indicators used to monitor how the scheme performed after rollout.
Tens of millions of claims · ¥10B+ fund · 20+ reports delivered to the Hunan Provincial Healthcare Security Administration
02
Personalized itinerary recommendation from online reviews and travelogues
Crawled multimodal review data from Qunar, cleaned and entity-aligned it, then modeled sentiment to build traveller profiles for itinerary recommendation.
Dataset of 5,000+ attractions and 7M+ reviews and travelogues · supported five SCI papers from the team
03
Multi-source intelligence fusion for emergency decision-making under big data
Built a hierarchical CNN–LSTM–MLP model that separates word–word relations from word–class relations, then validated it on 5,477 emergency news stories and 2,815 ordinary ones.
99.55% recognition and 94.82% classification accuracy
04
Big-data-driven adaptive traffic signal timing
Formulated a multi-objective program minimising delay while maximising throughput, solved it with a genetic algorithm, and validated the result on a multi-junction road network in VISSIM.
05
Education & Appointments
Laboratory Engineer, School of Artificial Intelligence
Wuhan University of Technology · Wuhan, China
Ph.D. in Management Science and Engineering
Central South University · School of Business, Department of Information Science
Advisor: Prof. Hongyu Zhang
Visiting Ph.D. Student, CSC-funded
National University of Singapore · Department of Data Science and Statistics
M.Eng. in Safety Science and Engineering
Wuhan University of Technology · School of Safety Science and Emergency Management
B.Eng. in Safety Engineering
Sichuan University of Science and Engineering
06
Honors
China Scholarship Council visiting fellowship
National University of Singapore
“Self-Improvement Star”
School of Safety Science and Emergency Management, WUT
Outstanding Graduate Student
Wuhan University of Technology
National Third Prize, 5th National Graduate Smart City Competition
National-level competition
National Scholarship
Ranked first in the cohort
Outstanding Student Cadre
Sichuan University of Science and Engineering
07
Contact
I am open to research collaboration in AI, health informatics, and information systems. Email is the fastest way to reach me.
lvlong@whut.edu.cn