삶의 질 향상을 위한 이미지 분할 기술의 발전과 응용 = Image Segmentation Advancements and Applications for Quality of Life Enhancement

다국어 초록 (Multilingual Abstract)

As the demand for precision and efficiency in medical diagnosis increases, image-based diagnostic technologies are gaining attention as a key means to improve quality of life. In particular, image segmentation, which enables precise identification of organs and lesions to assist diagnosis and treatment planning, is being actively studied for clinical applicability. This paper reviews the development of medical image segmentation techniques over the past decade and analyzes their use in real-world clinical settings. Segmentation approaches are categorized into three model families: convolutional neural networks (CNN), transformers, and foundational models. We examine each model family’s technical features and clinical use cases, and compare them in terms of implementation requirements and usability. These technologies are expected to support diagnosis, treatment planning, and intraoperative image analysis, ultimately contributing to improved quality of life.