< BK21 플러스 BEST 정보기술 사업단 세미나 개최 안내 >

개최일시 : 2019년 10월 22일 (화) 11:00 ~ 12:00

개최장소 : 제 4공학관 D509호

세미나 제목 : Quantitative PET Imaging Techniques in PET/CT and PET/MRI Systems

내용 :

Positron emission tomography (PET) is an in vivo functional imaging tool that measures physiological properties by using radiopharmaceuticals. However, PET images without compensation for physical effects (e.g, attenuation, scatter, respiratory motion, etc) are not quantitative, which should be corrected by quantitative imaging techniques to precisely estimate radiotracer concentration in the body. In terms of attenuation correction (AC), the introduction of combined PET/MRI scanner has grown research interest for PET AC in the past several years, since MRI data do not provide direct information for electron density. In terms of motion correction (MC), respiratory motion causes blurring artifacts which result in qualitative and quantitative errors such as tumor misquantification, misdisplacement, and volume overestimation.

Therefore, my research is focused on developing techniques that correct physical artifacts such as attenuation and respiratory motion to improve the accuracy of quantitative PET imaging, translating the techniques into conventional and emerging commercial systems for clinical practice.

In this talk, I will present (1) recent techniques in overcoming attenuation correction challenges in PET/MRI imaging, (2) CT-less AC through deep convolutional neural networks potentially applicable to both conventional and emerging PET and SPECT scanners, (3) practical approaches in removing motion blurring artifacts, in order to address currently unmet clinical needs in both PET/CT and PET/MRI.

강연자 성함&직함 / 소속 : 양재원 박사 / Associate Scientist, Department of Radiology & Biomedical Imaging, University of California - San Francisco

초청자 : 전기전자공학과 교수 김동현