Yechan Seo

I aspire to become a surgeon who expands the boundaries of medicine through biomedical engineering, with a focus on advancing medical robotics and neural interface technologies, ultimately driving growth in the medical device industry and contributing to academic advancements in the field.

See my work Yechan Seo

Professional Experience

  • Rosota Inc.
    CEO (January 2025 - Present) · Leader, Rosota Intelligence Institute (RII)
    Homepage : rosota.run · Rosota Intelligence Institute (RII)
  • Seoul National University Hospital, Medical Spatial Computing (MEDISC) Labs
    Researcher (December 2021 - Present)
    • Medical AI: Research on medical AI, including task-specific models and foundation models for clinical applications.
    • Medical Robotics: Conducting research on medical robotics development with focus on teleoperation-based imitation learning and reinforcement learning systems. Optimizing human-robot interaction (HRI) through BCI and XR technologies, with emphasis on data collection pipeline architecture and system design for autonomous task execution.
    • BCI Research: Investigating neural mechanisms through encoding (actuation) and decoding (biological signal interpretation) for practical applications.
    • MEDISC Researchers Directory
  • Laplacian
    Founder and Representative (January 2024 - December 2024)

    A research group dedicated to developing control systems for medical robotics leveraging XR and BCI technologies.
    Selected for the Seoul National University Center for Entrepreneurship & Innovation support program, receiving funding and mentorship.

  • AIRS Medical
    AI Research Scientist (January 2022 - December 2022)

    Conducted research on AI methods for MRI acceleration.

Publications & Conference Presentations

  • Co-Author: Drape-Compatible Tool-Tip Localization for Hand-Held Laparoscopic Instruments via UWB Carrier-Phase Ranging and Trocar-Constrained Geometry
    Submitted to IEEE ICRA 2027 · arXiv preprint (September 2026)
    https://arxiv.org/abs/2609.39129
    Research Summary: Tool-tip localization using UWB carrier-phase ranging, inertial sensing, and trocar constraints, designed to operate with sterile drapes.
  • Co-Author: Contact-Adaptive Robotic Ultrasound Probe Control for Tissue Exploration and Continuous Task-Relevant Visualization Using Robot-Free Image-Motion Demonstration
    Submitted to IEEE ICRA 2027 · arXiv preprint (September 2026)
    https://arxiv.org/abs/2609.37456
    Research Summary: Demonstration-informed ultrasound probe control with force regulation to maintain a clinician-selected view, evaluated on a dynamic bladder phantom.
  • Co-Author: Surgical Master Console Using General-Purpose Robot Arms and a Separable Articulated Distal Interface: Porcine In-Vivo Evaluation
    Submitted to IEEE ICRA 2027 · arXiv preprint (September 2026)
    https://arxiv.org/abs/2609.37455
    Research Summary: A bimanual surgical master console built from general-purpose robot arms and separable wrist-and-grasp interfaces, evaluated in a porcine procedure.
  • Co-Author: Contact-Stable Deformable Tissue Simulation Using Implicit Integration and Live-Pose Grasp Constraints for Laparoscopic Surgery Robot Policy Evaluation
    Submitted to IEEE ICRA 2027 · arXiv preprint (September 2026)
    https://arxiv.org/abs/2609.25642
    Research Summary: A reconstructed deformable-tissue simulation environment with stable contact and grasp constraints for closed-loop surgical robot policy screening.
  • Co-Author: From Instrument-Mounted Demonstrations to In-Vivo Execution: Learning Bimanual Laparoscopic Appendectomy Without Robot-Collected Demonstrations
    Submitted to IEEE ICRA 2027 · arXiv preprint (September 2026)
    https://arxiv.org/abs/2609.25625
    Research Summary: An end-to-end pipeline for learning bimanual surgical policies from hand-held instrument demonstrations, with simulation-based policy selection and in-vivo rabbit evaluation.
  • Co-Author: Recording Hand-Held Laparoscopic Instrument Motion in the Operating Room: Magnetometer-Free Fusion of Inertial, Range and Visual Sensing
    Submitted to IEEE ICRA 2027 · arXiv preprint (September 2026)
    https://arxiv.org/abs/2609.25577
    Research Summary: A clip-on instrument logger combining inertial, range, and visual sensing to recover laparoscopic instrument motion without magnetic measurements.
  • First Author: Data-Efficient Surgical Video Segmentation with Sparse Point Supervision
    Poster submitted to the SNU 80th Anniversary Convergence Healthcare Symposium (September 18, 2026)
    Symposium information
    Research Summary: Developed a data-efficient surgical video segmentation pipeline using sparse point annotations, SAM 2.1 temporal mask propagation, and YOLO11s-seg. Generated 59,588 masks from 292 annotated points; the reference model achieved mask mAP50 of 0.921 at 54 FPS. At comparable image budgets, broader clip coverage outperformed denser sampling within fewer clips. Evaluation used automatically generated labels from a single surgical session; independent manual-mask validation remains necessary.
  • First Author: Generating a Realistic Rolling Illusion in Immersive Virtual Reality Environment Using Galvanic Vestibular Stimulation
    Presented at the Fall Academic Conference of The Korean Society of Medical Informatics (KOSMI) (2024)
    Research Summary: Investigated the efficacy of transcranial alternating current stimulation (tACS) applied to the vestibular nerve for inducing a rolling illusion in a VR environment, enhancing the sense of presence.
  • Co-First Author: Machine-learning-based models for the optimization of post-cervical spinal laminoplasty outpatient follow-up schedules
    Published in BMC Medical Informatics and Decision Making, 24:278 (September 30, 2024)
    https://doi.org/10.1186/s12911-024-02693-y
    Research Summary: Developed and validated a machine learning-based framework to optimize clinical follow-up schedules for patients undergoing cervical laminoplasty. Leveraged time-series patient data, including the Japanese Orthopedic Association (JOA) scores from multiple postoperative timepoints, to predict 2-year outcomes using various machine learning algorithms. Achieved the highest predictive performance (AUC 0.90 ± 0.13) with an LSTM-based model.
  • First Author: Development of an EEG-Based Real-Time Attention State Classification Machine Learning Model: Enhancing Achievement in Medical VR Training
    Presented at the Korean Society of Computer Information (KSCI) Spring Academic Conference (2024)
    Award: Best Paper Award
    Research Summary: Developed a real-time EEG-based attention state classification model integrated into a VR training program for preoperative education, achieving a high performance (AUROC = 0.9336).
  • First Author: Functional Near-Infrared Spectroscopy Signals as Biomarkers for the Quantitative Assessment of Nausea: A Pilot Study
    Presented at the Spring Academic Conference of The Korean Society of Medical Informatics (KOSMI) (2024)
    Research Summary: Conducted a pilot study exploring the use of functional Near-Infrared Spectroscopy (fNIRS) signals as biomarkers for objectively quantifying nausea, a primary symptom of cybersickness. Achieved a significant predictive accuracy with an R² value of 0.9045.

Research Projects

  • Neuroadaptive Mixed Reality & BCI Robotics
    2025
    • Developing OpenGalea, an open-source neuroadaptive mixed reality system integrating BCI and XR using OpenBCI and Meta Quest 3.
    • Designed a brainwave-driven interaction system enabling real-time mental state classification (e.g., Attention vs. Relaxation) and co-located collaborative MR experiences.
    • Implemented a custom machine learning model for real-time EEG-based attention state classification, facilitating seamless BCI-controlled interactions.
    • Developed a multiplayer MR experience where users synchronize their brainwave states to control virtual environments.
    • Engineered an affordable EEG-integrated headset by modifying the OpenBCI Ultracortex IV, reducing costs by over 15x compared to commercial solutions.
    • Awarded "Best MR Lifestyle App for Meta Quest - Honorable Mention (Top 3)" at MIT Reality Hack 2025.
    • GitHub Repository | Devpost | Demo Video
  • AI & Robotics Research Expertise
    2024 - Present
    • Expertise in embodied AI and medical robotics, including Vision-Language-Action (VLA) models, world models, and BCI-XR interaction frameworks.
    • Research Focus: VLA-based policy learning, world modeling, imitation learning, and reinforcement learning for medical robot control and autonomous task execution, supported by teleoperation data collection pipelines and robotic system architecture.

Education

  • Seoul National University College of Medicine
    March 2021 - February 2028 (Expected Graduation)
    Major: Medicine
    • Premedicine Program: March 2021 - February 2023 (Completed)
    • Medicine Program: February 2023 - February 2028 (Expected Graduation)
  • Gap Year for AI, BCI, Robotics, and XR Research (2024)
    Conducted interdisciplinary research on AI, BCI, robotics, and XR at the Department of Transdisciplinary Medicine, Seoul National University Hospital.
    Advisor: Prof. Hyoun-Joong Kong
    Research Lab: https://snuh.medisc.org/
  • Gyeonggi Science High School for the Gifted
    Graduated: February 2021
    Major: Physics and Mathematics (Valedictorian)
  • Curriculum Path Model (CPM)
    Completed a specialized curriculum within the Premedicine program at Seoul National University College of Medicine.
    • Linear Algebra 1, 2
    • Introduction to Analysis
    • Concepts and Practices of Computers
    • Statistics for Life Sciences
    • Mathematical Foundations of Deep Neural Networks
    • Introduction to Data Science
    • Venture Startup and Web Programming 2
    • Electricity and Magnetism
    • Quantum Physics 1, 2

Awards and Honors

  • Hwang Jun-Sik Student Fellowship (2025, 2024, 2023)
  • Best MR Lifestyle App for Meta Quest - Honorable Mention (Top 3), MIT Reality Hack (2025)
  • Seoul National University College of Medicine Future Talent Award (Kwon Jun-Soo Clinical Medicine Student Research Award) (2024)
  • Best Paper Award, Korean Society of Computer Information (2024)
  • 2nd Place, Seoul National University Hospital LLM Hackathon (2024)
  • 4th Place, Bundang Seoul National University Hospital Datathon (2023)
  • Student Research Award, Seoul National University College of Medicine Premedicine Program (2022)
  • Priority Selection, Korean Physics Olympiad Winter School (2019, 2020)
  • Honorable Mention, International Mathematical Modeling Challenge (IMMC) 2019
    Represented South Korea as a national team member.

Technical Skills

  • Programming: Python, C++, MATLAB
  • Machine Learning Frameworks: PyTorch, TensorFlow
  • AI and LLM Applications: LangChain, Hugging Face Transformers
  • Biosignal Analysis: fNIRS, EEG, EMG (preprocessing, feature extraction, and classification)
  • Data Analysis: Pandas, NumPy, SciPy
  • Robotics & Embodied AI: VLA Models, World Models, Imitation Learning, Reinforcement Learning, Isaac Sim, ROS / ROS 2, Control Algorithms
  • Real-Time Systems: Pylsl (real-time bio-signal communication and ML inferencing systems)

Leadership Experience

  • IT Director, Seoul National University College of Medicine Student Council
    March 2022 - February 2023, December 2023 - Present
    • Managed the student council's web server and drive system
    • Oversaw IT infrastructure
  • MD Winners 18th President, Seoul National University College of Medicine
    18th President: 2024.01.~2024.12., 17th Vice President: 2023.01.~2023.12.
    MD Winners, founded in 2007, is the first medical economics and business society in South Korea, affiliated with Seoul National University College of Medicine. The society aims to foster future medical entrepreneurs, policymakers, and healthcare managers under the vision of "Growing the Pie" to drive growth in Korea's healthcare industry. MD Winners consists of 116 members, including an 8-member executive team, dedicated to nurturing excellence in medicine and beyond. Key Events Organized by MD Winners:
    • Healthcare Management Symposium (HMS): An annual conference showcasing team project results and featuring expert speakers from the healthcare industry, finance, and academia. Successfully hosted the 2024 HMS, a medical startup-focused event, attracting 150 medical students nationwide.
    • Winners Talk Seminar Series: Biweekly seminars fostering practical discussions with professionals from various fields to sustain interest in healthcare entrepreneurship.
    • Hackathons: Regularly held hackathons during semester breaks, focusing on ideating and developing MVPs for practical healthcare solutions. In 2024, organized two AI/XR hackathons in collaboration with the Department of Transdisciplinary Medicine at Seoul National University.
    • Team Projects: Year-long team-based research and development projects aimed at creating impactful outputs in the medical industry.
    • For more details, visit the official MD Winners website.
  • Medical Frontiers Cooperation (MFC) 57th President (United Medical Volunteer Group)
    A joint medical volunteer group comprising Seoul National University College of Medicine, Seoul National University School of Dentistry, and Ewha Womans University College of Medicine.
    The group organizes medical volunteer activities such as seasonal clinics in rural areas and weekend clinics in urban regions to support underserved communities.
    • Conducted seasonal medical volunteer activities in Gyechon-ri, Pyeongchang, Korea
    • Hosted weekend clinics in Jongno-gu, Seoul, Korea
  • GIV (Global Inter-culturing and Volunteering) 37th President
    2022 Fall Semester (Executive Team: 2021 Fall Semester - 2022 Fall Semester)
    • Led a volunteer delegation to Taipei Korean School in Taiwan (2023.02.05 - 2023.02.15)
    • Organized educational volunteering trips to Yongdae Elementary School in Inje, Korea:
      • 1st Trip: 2022.07.31 - 2022.08.06
      • 2nd Trip: 2024.08.04 - 2024.08.10
  • FOS (Fist of SNU)
    Captain of Seoul National University Boxing Club (2022)

Languages

  • Korean (Native)
  • English (Professional Proficiency)

News & Media

Coverage of Yechan Seo and Rosota Inc., grouped by topic.

RESEARCH & DEVELOPMENT6 ARTICLES
BUILDING ROSOTA10 ARTICLES
BEYOND BORDERS1 ARTICLE
MEDICINE & COMMUNITY1 ARTICLE
CLINICAL AI RESEARCH8 ARTICLES
ACADEMIC MILESTONES3 ARTICLES

University & Research Lab Updates

Contact Me