몽발개발

Deep learning methods and applications in neuroimaging 본문

뇌공학/논문 정리

Deep learning methods and applications in neuroimaging

집사 몽이 2021. 1. 15. 14:27
반응형
읽은 날 2021.01.15. 학술지 Journal of Neuroscience Methods
제목 Deep learning methods and applications in neuroimaging
저자 Jing Sui, MingXia Liu, Jong-Hwan Lee, Jun Zhang, Vince Calhoun
한줄요약 Neuroimaging에 쓰이는 MLDL기법들의 기존 한계점과 그를 극복한 과정
초록 Deep learning (DL) has gained considerable attention in the scientific
community, breaking benchmark records in many areas such as
speech and visual recognition. However, the incorporation of deep
learning approaches in neuroimaging is still a challenging and promising
direction, due to the high-dimensional feature dimensions and
limited sample sizes (Calhoun and Sui, 2016). MRI features. Currently,
advances in medical imaging technologies have enabled image acquisition
at faster rates and with increased resolution. Also, multiple accessible
international brain imaging datasets online facilitate the generation
of neuroimaging big data. These provide wonderful testbeds for
the advanced computerized tools, especially deep learning approaches,
which has shown its efficacy to neuroimaging applications (Hou et al.,
2019; Kim et al., 2016; Liu et al., 2018; Yan et al., 2019).
키워드 Neuroimaging, fMRI, Machine learning, Deep learning, CNN, RNN
의의 fMRI등 뇌영상을 통해 특정 disorder의 biomarker를 찾기 위한 여정이 정리됨.
 비판점  

 

반응형