Yi-Chen Sun Academic Website
Research Vision

Translating fundamental medical AI into clinically resilient, real-world deployment.

I am a direct-entry Ph.D. student at National Yang Ming Chiao Tung University (NYCU) focused on Health Research. My work bridges the gap between clinically-aligned, rigorous algorithm development—such as my proposed DBI-MambaUNet and Size-Aware Focal Tversky Loss for micro-lesion sensitivity—and the stringent demands of actual medical workflows.

Rather than stopping at theoretical performance, I build end-to-end systems. My research spans contrast-free brain tumor MRI analysis via knowledge distillation, the development of 3D medical vision-language models (VLM) for automated clinical report generation, hallucination-resistant structured reporting natively integrated with clinical PACS, and leading teams to scale TFDA/FDA-facing oncology systems.

Publications Projects CV Contact
AOCR Kaggle Champion
AOCR (Top-3 Radiology Conference) Kaggle AI Champion, National Innovation Award Recipient, and featured in an exclusive interview.
CTO & Startup
Served as startup CTO, participating in the FITI program and selected as a Model Success Case by MOE Taiwan.
StanCode Instructor
Served as an Instructor at the StanCode Programming Education Institute and acted as an AI Poster Mentor, leading student teams in hands-on AI project deployments.
TFDA & U.S. FDA
Contributed to Brain Tumor SaMD operations, achieving both TFDA and U.S. FDA clearance for AItewan DeepBT Detector–Plus.
GPA: 4.11 / 4.30
Achieved a near-perfect GPA in NYCU's highly competitive Direct-Entry Ph.D. program, balancing rigorous coursework with pioneering medical AI research.
7 Publications
Authored 3 journal papers (with 1st-author at MedIA IF=11.8 under review) and 4 accepted conference papers (all Oral Presentations).

Honors

Selected Awards

  • Champion, AOCR Kaggle AI Challenge, held under the Asian Oceanian Congress of Radiology (AOCR), one of the three major international radiology conferences.
  • 21st National Innovation Award (Academic Innovation Award), Taiwan.
  • Served as startup CTO, participating in the FITI program and selected as a Model Success Case by MOE Taiwan.
  • Winner, AI Application Competition, Digital Industry Administration, Taiwan.
  • Shortlisted for the Smart Innovation Award.
  • Served as an Instructor and AI Poster Mentor at the StanCode Programming Education Institute, leading student teams in hands-on AI project deployments.

Translation

Innovation

  • Co-developed AItewan DeepBT Detector–Plus, with both TFDA and U.S. FDA medical device approvals obtained.
  • Independently developed AItewan DeepBT Detector–A Plus, with TFDA and U.S. FDA medical device applications pending.

IP

Patents

  • Implementation Method and Electronic Device for Acute Appendicitis Assisted Diagnosis: Taiwan utility patent granted; Taiwan invention patent pending.
  • System and Method for Multimodal Language Models in Medical Imaging Diagnostic Assistance: Taiwan and U.S. invention patents pending.

Research Interests

Current Focus

Health Research: Bridging fundamental medical image analysis, multimodal vision-language systems, and rigorous clinical translational workflows to deploy trustworthy AI in oncology and radiology.
Current Focus: Advancing Medical AI by translating lesion-sensitive segmentation to clinical use, bridging the gap between algorithm development and real-world clinical deployment.
Technical Core: 3D MRI brain tumor segmentation, Non-contrast MRI analysis using knowledge distillation, and structurally constrained VLM-based report generation.
Academic Rigor: Submitted a first-author fundamental manuscript to Medical Image Analysis (IF=11.8) during the first year of a direct-entry Ph.D. program, alongside multiple accepted Oral Presentations.
Translational Scale: Led the development of end-to-end AI systems that have already successfully navigated Taiwan FDA (TFDA) and U.S. FDA regulatory pipelines, currently deployed in top-tier medical centers.

Why This Profile Stands Out

Selected Advantages

Bench-to-Bedside Translation

Deeply committed to Health Research that scales. My systems do not stop at paper publication; they are built for regulatory clearance (TFDA/FDA) and active integration into live medical workflows.

Methodological Rigor

Innovating at the foundation of Medical AI: developing DBI-MambaUNet for micro-lesion sensitivity and creating contrast-free inference frameworks using advanced knowledge distillation.

Recognized Leadership

Demonstrated capacity to orchestrate health tech projects—ranging from serving as CTO of an AI rehab platform to winning the prestigious AOCR Kaggle AI Challenge.

Multimodal Clinical Synthesis

Pioneering end-to-end integration that goes beyond imaging masks to generate clinically sound, hallucination-resistant structured radiology reports via constrained LLMs.

Selected Projects

Featured Work

A curated set of the most representative projects from research, mentoring, challenge leadership, and translational AI.

National Yang Ming Chiao Tung University, Institute of Biophotonics
Ph.D. Researcher 2024–Present

End-to-End Non-Contrast Brain Tumor SRS Clinical AI System

Flagship Research

An integrated research program centered on brain tumor stereotactic radiosurgery. It features an advanced pipeline combining lesion-sensitive segmentation, quantitative MRI analysis, and constrained VLM-based report generation—all enabled on non-contrast MRI via knowledge distillation.

National Yang Ming Chiao Tung University, Institute of Biophotonics
Ph.D. Researcher 2024–Present

DBI-MambaUNet + SA-FTL for Micro-Lesion Segmentation

Method Development

A dedicated method-development project for multi-subtype brain tumor micro-lesion analysis in stereotactic radiosurgery, with explicit focus on improving lesion sensitivity for small lesions.

AItewan BioMedical Technology Inc.
Senior AI Engineer & Product Manager Ongoing

AItewan DeepBT Detector–Plus

Medical AI Product

A clinically oriented brain tumor AI product successfully transitioning from research to clinical validation. It has achieved both Taiwan FDA (TFDA) and U.S. FDA (510(k)) clearance. It is now officially deployed and in actual clinical use at Taipei Veterans General Hospital, Taichung Veterans General Hospital, and Shin Kong Wu Ho-Su Memorial Hospital.

AItewan BioMedical Technology Inc.
Senior AI Engineer & Product Manager Ongoing

AItewan DeepBT Detector–A Plus

Medical AI Product

An advanced brain tumor AI product driving the next generation of diagnostics. Strongly positioned for both Taiwan FDA (TFDA) and U.S. FDA medical device applications. Like its predecessor, this system has already been deployed for clinical use at Taipei Veterans General Hospital, Taichung Veterans General Hospital, and Shin Kong Wu Ho-Su Memorial Hospital.

NOVATERA (ICF Smart AI Rehabilitation System)
Chief Technology Officer (CTO) 2025–2026

NOVATERA: ICF Smart AI Rehabilitation System

Innovation / Entrepreneurship

An AI-assisted rehabilitation platform designed for home and clinical environments, developed under CTO leadership with a focus on functional assessment, workflow design, and translational deployment.

View Project
AOCR Kaggle AI Challenge Project
Research Team Leader 2023–2024

AOCR Kaggle AI Challenge: Acute Appendicitis CT Segmentation

Challenge Champion

A 3D CT segmentation framework for acute appendicitis built for the AOCR Kaggle AI Challenge, emphasizing lesion-sensitive detection, localization, and clinically robust performance.

Watch Interview National Innovation Award
StanCode (founded by a Stanford alumnus, John Stephens Jr. Memorial Award recipient)
AI Poster Mentor 2025–2026

Medication Delivery Robot Mentoring Project

Mentoring

A healthcare-oriented student project focused on a smart medication delivery robot, combining computer vision, sensor fusion, autonomous navigation, and practical system design thinking.

YouTube Instagram
StanCode (founded by a Stanford alumnus, John Stephens Jr. Memorial Award recipient)
AI Poster Mentor 2025–2026

Medication Recognition App with RAG + LLM

Mentoring

A mentoring project on medication recognition and grounded medical education using image detection/classification, retrieval-augmented generation, and large language models.

YouTube Instagram