|
Tianzuo Yuan
Email: maxyuan0622@outlook.com | cc31642@um.edu.mo | cs25966@bristol.ac.uk Research Interests
Biomedical Imaging & Engineering • Computer Vision, Deep Learning & Machine Learning • Artificial Intelligence and Multi-agent LLMs • Multimodal Fusion & Bioinformatics Algorithms
|
I am an undergraduate student at the University of Macau and a member of the Honours College, pursuing a BSc in Bioinformatics with a GPA of 3.48/4.0 (Top 15%). My research focuses on the intersection of Artificial Intelligence and Medical Science, driven by a commitment to solving real-world clinical challenges through data-driven methodologies.
My expertise is grounded in multiple scientific projects centered on Early Cancer Diagnostics. I specialize in Multimodal Fusion, Attention Mechanisms, and YOLO-based detection, with practical applications in lung nodule detection, brain tumor segmentation, and real-time endoscopic analysis. Beyond healthcare, I have successfully applied deep learning to diverse fields such as autonomous driving, remote sensing, and financial sentiment analysis, demonstrating a strong capability for algorithmic adaptation across complex data types.
Currently, I am exploring frontier generative AI, specifically Diffusion Models for bi-directional generation between genetic and imaging data. In addition to algorithm design, I possess the full-stack capability to architect and deploy integrated platforms and Multi-agent collaborative systems, bridging the gap between theoretical research and practical deployment.
I am driven by the vision of contributing to the field of Medical AI by transforming advanced methodologies into deployable, high-impact tools that improve clinical outcomes and patient care. To this end, I am currently advancing my latest research and preparing submissions for MICCAI and NeurIPS 2026.
Education
- University of Macau (Macau, China) — BSc in Bioinformatics — 08/2023–08/2027 — GPA: 3.48/4.0 (Top 15%), Honours College GPA: 3.77/4.0
- University of California, Berkeley (CA, U.S.A.) — Visiting Student, Summer Program — 06/2024–08/2024 — Electives: Human Anatomy, Statistics & Probability, Cognitive Neuroscience
- University of Bristol (Bristol, U.K.) — Exchange Student — 01/2026–06/2026 — Electives: Data Science Perspectives, Evolutionary Genomics, Recombinant DNA Technology
Selected Publications
|
Intelligent Detection Model of Lung Nodules in Medical CT Images
7th International Conference on Electronics and Communication, Network and Computer Technology (ECNCT), Guangzhou, 2025. doi: 10.1109/ECNCT66493.2025.11172473 Best Oral Presentation Award • Excellent Young Scholar Award
|
|
A Novel Adaptive Superb Fairy-Wren Optimization Algorithm for Numerical Optimization Problems
Biomimetics (Q1, IF: 3.9), 2025. doi: 10.3390/biomimetics10080496
|
|
Multimodal Deep Learning Framework for Brain Tumor Segmentation Using CT and MRI Images
4th International Conference on Image Processing, Computer Vision and Machine Learning (ICICML), Chongqing, 2025. doi: 10.1109/ICICML67980.2025.11333677
|
|
MRI-based Artificial Intelligence in Predicting Prostate Cancer Biochemical Recurrence: A Systematic Review and Meta-Analysis
Journal of Medical Internet Research (JMIR), under revision
|
|
A Lightweight, Deployable AI Agent for Cancer Care Navigation: A Data-Driven Approach for Regions with Unique Epidemiological Profiles
The 14th IEEE International Conference on Healthcare Informatics (ICHI 2026), under review
|
|
Adaptive Fusion Multi-Agent System for Gastric Cancer Detection: Theoretical Analysis with Excellent Cross-Domain Generalization
Biomedical Signal Processing and Control (BSPC), under review
|
Research Experience
|
Pan-NEN Transcriptomic Atlas & AI Diagnostic System (NETA)
Established a comprehensive Pan-Neuroendocrine Tumor (Pan-NEN) Transcriptomic Atlas integrating 308 high-quality RNA-seq samples from 9 independent GEO datasets, covering NENs from four major organ sites (Lung, Pancreas, Prostate, Skin). Identified 3 robust molecular subtypes (C1: Immune-Hot, C2: Metabolic, C3: Neuronal) using Consensus Clustering. Developed an XGBoost-based AI diagnostic model with 100% training accuracy, featuring 50 signature genes selected via Boruta algorithm and SHAP interpretability. Characterized immune landscape via ssGSEA (28 immune cell types), predicted drug sensitivity, and validated prognostic markers using external TCGA-PCPG data. Built an interactive R Shiny web platform (NETA-Web) with modules for gene expression analysis, immune landscape profiling, drug sensitivity prediction, survival analysis, and AI-powered subtype diagnosis.
|
|
A Cross Deep Learning System for Rapid Automatic Diagnosis of Gastric Cancer from Real-Time Endoscopic Videos
Developed deep learning pipeline for early gastric cancer diagnosis from real-time endoscopic video, achieving high accuracy for stage detection. Developed hybrid AI model using VGG16 and XGBoost for multi-stage cancer detection, achieving high classification accuracy. Developed multimodal AI models combining imaging and clinical data for early cancer diagnosis, focusing on improving predictive performance with deep learning (05/2025–11/2025).
|
|
CavengerX: Ultrasound-Activated Calcium Overload Anti-Cancer System
Developed a synthetic biology-based cancer therapy system that induces calcium ion overload in cancer cells through ultrasound activation. Contributed to conceptualization, project administration, software development, and wiki design. Designed and implemented homepage cover, loading interface, progress bar animations, parallax scrolling effects, and dynamic visualizations for cancer cell lysis. Built interactive web platform using HTML, CSS, and JavaScript with Bootstrap framework. Project wiki: 2024.igem.wiki/um-macau/
|
Internship Experience
|
Multi-Agent Systems & AI Framework Design
Contributed to design and implementation of multi-agent AI framework for smart cities, integrating NLP and decision-making systems. Developed a domain-specific language (DSL) framework for multi-agent coordination with React front-end and Python FastAPI backend, supporting 1000+ agents with <200ms response time. GitHub: github.com/Max-YUAN-22/Agent_DSL
|
|
Single Cell Data Analysis & Bioinformatics
Conducted RNA-seq and single-cell transcriptomic data analysis, including differential expression and cell clustering. Applied tools like Seurat and Scanpy for data preprocessing, visualization, and interpretation.
|
Awards & Honors
- Future Cup - 4th University Big Data Challenge (2025) [Second Prize in Undergraduate Group]
- Bronze Medal, Kaggle RSNA Intracranial Aneurysm Detection, 2025
- iGEM - Silver Medal (2024) CavengerX: Ultrasound-Activated Calcium Overload Anti-Cancer System
- Best Oral Presentation, ECNCT 2025
- Excellent Young Scholar Award, ECNCT 2025
- Dean's Honor List, University of Macau, 2024–2025
- Honours College Certificate Program, University of Macau, 2024–2025
- Asia-Pacific Mathematical Contest in Modeling (2024) [Outstanding Participation Award for underwater image enhancement research]
- Participated in Mathematical Contest in Modeling (MCM, 2025) and various Kaggle competitions
Skills
- Programming: Python, C++, MySQL, R, SPSS, MATLAB
- Deep Learning Frameworks: PyTorch, TensorFlow, Keras
- Tools: Git, Linux, Docker, Jupyter Notebooks
- Languages: English (TOEFL iBT 98), Mandarin, Cantonese (Basic)
Extracurricular Activities
- 8/2025–12/2025: Member of the Unmanned Technology Application Team, guided and assisted lower-grade undergraduate students in participating in computer vision and AI-related projects.
- 6/2024–9/2025: Member of the Campus Stray Cat Rescue Program, developed mini-programs and apps to protect the safety of stray cats and record feeding, designed cultural and creative products, and organized a series of visits to pet stations and community care services on behalf of CKLC college of the University of Macau.
- 08/2024–12/2025: Minister of the Publicity Department, IET Hong Kong Branch Youth Membership for the Macau Student Organization.
- 11/2023–11/2024: Member of the Publicity Department, FHSSA, Faculty of Health Sciences, University of Macau.