Department of Convergence Medicine
University of Ulsan College of Medicine
Asan Medical Center
88 Olympic-Ro 43-Gil Songpa-Gu
Seoul 138-736, South Korea
The medical imaging and intelligent reality lab. (MI2RL) has been established at the department of convergence medicine and radiology, University of Ulsan College of Medicine (UoU), Asan Medical Center (AMC), which is one of the leading hospitals in South Korea. The MI2RL is directed by many medical doctors and Prof. Namkug Kim who received his BS, MS, and Ph.D. degrees from the Department of Industrial Engineering at Seoul National University and is the author of about 160 peer-reviewed original articles (5344 citations, 40 h-ind, 101 i10-ind) and 100 patents. We currently have about 50 lab members with various backgrounds including medical doctors, engineers, scientists, radiographers, and nurses (Fig. 1) and we jointly work with many other surgeons in Seoul National University Hospital and Yonsei Severance Hospital, etc.
The motivation of the MI2RL is to develop and apply medical image based radiological and surgical applications including deep learning based computer-aided detection (CAD), computer aided surgery (CAS), robotics, 3d printing in surgery (3DPS), medical image processing (MIP), etc. in a real clinical environment (Fig. 2). The research areas are three-fold: (1) medical imaging understanding with MIP and deep learning; (2) CAS including robotics, stereovision, gesture recognition, patient monitoring, AR/VR, etc. Specifically, needle biopsy/RF ablation, knee replacement, flexible catheterization, congenital heart surgery, breast conversation surgery, orbital fractures, etc. have been studied; (3) 3DPS that combine deep learning, conventional image processing, augmented reality (AR), 3D printing technology, etc. The main research topics are further detailed below.
Robot for needle biopsy/RF ablation
Remote controlled CT-fluoroscopy needle biopsy and RF ablation robot has been developed and cleared at Korean FDA for human testing (Fig. 3). Image-based needle trajectory planning, virtual wall, and registration between a CT-fluoroscopy image and diagnostic CT images were also developed.
Various medical devices for interventional robotics
The MI2RL developed various medical devices for interventional robotics including stereovision, various kinds of image based surgery/intervention planners, gesture recognition interface, patient monitoring with Kinect v2.0, AR/VR, and 3D printing phantoms for simulation and testing (Fig. 4).
3D printing for surgery
We evaluated more than 60 applications with 3D printing for surgery including breast conservation surgery, congenital heart surgery, orbital fracture reconstruction surgery, etc., which could be categorized into a patients-specific simulator, guide, and implants (Fig. 5). In addition, digital epitheses and imaging and experimental phantom have been developed.
Recently, we have studied surgical video understanding to recognize the steps of a surgical procedure, surgical device, and anatomical landmarks, etc., which could be augmented on video and/or endoscopy with AR for simple mastoidectomy for the first time and received positive feedback. Several technologies for vascular measurement methods in interventional imaging, 3DPM, and lung imaging quantification have been transferred to medical companies in Korea.
To see more on the MI2RL activity, please visit http://mirl.ulsan.ac.kr/