About Me
At a Glance
I am a postgraduate student with experience in medical image analysis. My academic journey began with computer vision and object detection during my undergraduate studies, and I am now expanding toward embodied intelligence and neuroscience-related topics such as brain-computer interfaces (BCI) and computational approaches to consciousness.
Education: M.S. in AI & Adaptive Systems (Distinction, Overall 79), B.S. in Computer Science
Learning Focus: Medical AI, Computer Vision, Embodied Intelligence, BCI (EEG/fMRI), Consciousness/DoC
Experience: Kaggle competitions (Top 4% Podcast Listening Time, Top 5% Calorie Expenditure, Top 7% PhysioNet ECG Digitization (Bronze)), Challenge Cup competition, National-level Undergraduate Innovation Training Program, 4 software copyrights, 1 patent application
Research Learning and Participation
Machine Learning Identifies Exosome Features Related to Hepatocellular Carcinoma
Frontiers in Cell and Developmental Biology (2022) · DOI: 10.3389/fcell.2022.1020415
Co-first author (third position): Designed the ML analysis pipeline; compared Random Forest, SVM‑RFE, and LASSO to identify and validate high‑value exosome biomarkers.
Multi-omics and Machine Learning-driven CD8+ T Cell Heterogeneity Score for Prognosis
Molecular Therapy Nucleic Acids (2024) · DOI: 10.1016/j.omtn.2024.102413
Provided ML support including LASSO regression to identify key prognostic genes from multi‑omics data and supply features for the CD8+ T cell heterogeneity score.
Using Multiomics and Machine Learning: Insights into Improving the Outcomes of Clear Cell Renal Cell Carcinoma via the SRD5A3-AS1/hsa-let-7e-5p/RRM2 Axis
ACS Omega (June 2025) · DOI: 10.1021/acsomega.5c01337
Implemented the complete ML analysis pipeline to identify and quantify the prognostic value of the SRD5A3‑AS1/hsa‑let‑7e‑5p/RRM2 axis in ccRCC, and contributed to validation using single‑cell and spatial transcriptomics.
## Partial Project Experience
Scientific Content Integrity & Review Platform (NDA)
Joint project under NDA: end-to-end pipeline for scientific PDF/image ingestion, content extraction, similarity/dedup checks, quality assessment, and integrity screening (incl. segmentation-based forgery cues), with review-oriented reporting
Python | PyMuPDF | MinIO | Elasticsearch | CV Similarity | Segmentation
Multi-Agent Document Workflow Platform (NDA)
Internship project under NDA: multi-role agent workflows for document processing; delivered a review system with OnlyOffice preview, OCR/parsing services, and offline-friendly Docker Compose deployment
FastAPI | React/Vite | Docker Compose | OnlyOffice | Multi-agent workflows | RAG | MCP
Multi-phase CT + Clinical Data Fusion
Multimodal modeling with multi-phase CT imaging and structured clinical variables for diagnosis/prognosis tasks, focusing on robust feature fusion and interpretability
PyTorch | Medical Imaging | Multi-modal Fusion | XAI
Partial Competitions and Awards
Predict Podcast Listening Time Competition
Ranked 116th/3310 (Top 4%) | Kaggle Global Competition
Predict Calorie Expenditure Competition
Ranked 178th/4316 (Top 5%) | Kaggle Global Competition
PhysioNet - Digitization of ECG Images
Bronze Medal | Ranked 97th/1424 (Top 7%) | Kaggle Global Competition
Challenge Cup Competition
Bronze Award | Zhejiang Province
National-level Undergraduate Innovation Training Program
National-level student research & innovation program