6/11 Image-driven precision oncology for malignant gliomas: beyond cliché (최승원 교수/삼성서울병원 신경외과)

작성자
kaistsoftware
작성일
2020-06-09 10:27
조회
11862
  • 강사 : 최승원 교수 (삼성서울병원 신경외과) 
  • 일시 : 2020. 6. 11 (목) 17:00~18:30
Malignant gliomas are most common primary brain malignancy in adults. Despite some success stories of other solid tumors, the cure for these tumors is still pessimistic. One of major hurdles to conquer these tumors is their rarity, and another concern is their peculiar anatomic location which may hinder adequate sampling. Given enormous intra-tumor heterogeneity, molecular biomarkers are limited to fully represent the biologic activities of the entire tumor as they are usually derived from a single confined sub-region of the tumor. During past decades, genomic research much helped to comprehend tumor biology of malignant gliomas, however, failed to improve the clinical outcome regarding precision oncology. In other words, genomic information alone could not recapitulate the tumor evolution underneath the clinical progression, which implying the necessity of integrating multi-omics data.
In the light of clinical application, medical images are powerful tools and potentially represent underlying biologic activity of tumors. Imaging biomarkers are more versatile than molecular biomarkers considering several key challenges against malignant gliomas; they are non-invasive and free from sampling bias, thus reflecting the characteristics of the whole tumor. However, they also have some inherent weakness; data reproducibility and objectivity are main concerns accompanied by radiographic assessment. Radiomics, by converting medical images into high-dimensional data, has enabled us to transform the descriptive imaging features into the format of quantitative and objective data, thus contributing the evolution of image-driven data science. Another unsolved issue is their translational ability in terms of biological activity. Many have attempted to annotate the imaging phenotype with understandable molecular correlates, still robust association has not been elucidated.
Herein we present the basic principles of radiomics and introduce several key radiomic studies in the field of neuro-oncology. Future perspectives of radiomics study will be also discussed in this session.
전체 143
번호 제목 작성자 작성일 추천 조회
공지사항
2025년 봄학기 콜로퀴엄 일정 안내
kaistsoftware | 2025.02.27 | 추천 0 | 조회 10646
kaistsoftware 2025.02.27 0 10646
142
6/2 Intelligent Techniques for Graphics, Vision, and Robotics (윤성의 교수/KAIST 전산학부)
kaistsoftware | 2025.06.02 | 추천 0 | 조회 460
kaistsoftware 2025.06.02 0 460
141
5/26 Denoising Diffusion for 3D Human and Object Pose Estimation Under Interactions (김태균 교수/KAIST 전산학부)
kaistsoftware | 2025.05.23 | 추천 0 | 조회 587
kaistsoftware 2025.05.23 0 587
140
5/19 이미지/비디오 생성 기술의 현재와 미래 (성민혁 교수/KAIST 전산학부)
kaistsoftware | 2025.05.16 | 추천 0 | 조회 814
kaistsoftware 2025.05.16 0 814
139
5/12 Startup Funding (최원호 교수/KAIST 전산학부)
kaistsoftware | 2025.05.08 | 추천 0 | 조회 1470
kaistsoftware 2025.05.08 0 1470
138
4/21 반복되는 SW오류, 어떻게 막을것인가? (허기홍 교수/KAIST 전산학부)
kaistsoftware | 2025.04.07 | 추천 0 | 조회 3382
kaistsoftware 2025.04.07 0 3382
137
4/7 Hacking Unmanned Vehicles (김용대 교수/KAIST 전기및전자공학부)
kaistsoftware | 2025.04.04 | 추천 0 | 조회 2845
kaistsoftware 2025.04.04 0 2845
136
3/24 Mobile AI Agent (신인식 교수/KAIST 전산학부)
kaistsoftware | 2025.03.21 | 추천 0 | 조회 3674
kaistsoftware 2025.03.21 0 3674
135
3/17 Analyzing LLM Inference Chains (유신 교수/KAIST 전산학부)
kaistsoftware | 2025.03.10 | 추천 0 | 조회 4047
kaistsoftware 2025.03.10 0 4047
134
3/10 AI 의인화와 윤리적 문제: AI는 어떻게 사람처럼 보이도록 설계되었는가? (김진형 교수/KAIST 전산학부)
kaistsoftware | 2025.03.05 | 추천 0 | 조회 6184
kaistsoftware 2025.03.05 0 6184
133
11/25 Finding Security Vulnerabilities in Layer-1 and Layer-2 Blockchains (강민석 교수/KAIST 전산학부)
kaistsoftware | 2024.11.21 | 추천 0 | 조회 7337
kaistsoftware 2024.11.21 0 7337