Kyungmu Lee
Seoul National University · 情報科学
研究室紹介
Professor Kyungmu Lee's research lab specializes in medical image analysis and computer vision, with a focus on developing deep learning and computational methods for enhancing diagnostic accuracy in medical imaging. Key research directions include automatic image enhancement for retinal and ultrasound images, interactive and weakly supervised segmentation, and 3D surface reconstruction from 2D images using shape-from-shading and shape-from-texture techniques. The lab emphasizes real-world clinical applicability by addressing challenges such as low-quality images, limited annotations, and user interaction efficiency.
Research Overview
Research Output Trend
Figures are computed from collected data and may differ slightly.
Selected Papers
15In this paper, we propose an automatic seed generation technique with deep reinforcement learning to solve the interactive segmentation problem. One of the main issues of the interactive segmentation problem is robust and consistent object extraction with less human effort. Most of the existing algorithms highly depend on the distribution of inputs, which differs from one user to another and hence need sequential user interactions to achieve adequate performance. In our system, when a user first
OBJECTIVES: To develop a convolutional neural network system to jointly segment and classify a hepatic lesion selected by user clicks in ultrasound images. METHODS: In total, 4309 anonymized ultrasound images of 3873 patients with hepatic cyst (n = 1214), hemangioma (n = 1220), metastasis (n = 1001), or hepatocellular carcinoma (HCC) (n = 874) were collected and annotated. The images were divided into 3909 training and 400 test images. Our network is composed of one shared encoder and two infere
In previous research (Tech. Rep. 172, University of Southern California, Los Angeles, Calif., 1991) we developed an iterative shape-from-shading (SFS) algorithm that uses a single image, by combining a triangular-element surface model with a linearized reflectance map. In the current research we generalize the single-image SFS algorithm to the photometric stereo SFS algorithm, which uses multiple images taken under different lighting condition for more-accurate surface reconstruction. An explici
PROBLEM: Low-quality fundus images with complex degredation can cause costly re-examinations of patients or inaccurate clinical diagnosis. AIM: This study aims to create an automatic fundus macular image enhancement framework to improve low-quality fundus images and remove complex image degradation. METHOD: We propose a new deep learning-based model that automatically enhances low-quality retinal fundus images that suffer from complex degradation. We collected a dataset, comprising 1068 pairs of