UCSF Center for Intelligent Imaging

Not every step of MRI reconstruction needs high precision. A custom CUDA kernel from @UCSFimaging #postdoc Nikhil Deveshwar stores k-space in 8-bit & runs the FFT in 16-bit on chip, about 3.5x faster than NVIDIA's 32-bit FFT with 4x less memory traffic. intelligentimaging.ucsf.edu/news/deep-gene…

If your imaging research runs on XNAT, here's your chance to see where the platform is headed. XNAT 2 enriches traditional research tasks with agentic AI & extends XNAT's reach into the clinic. Hear from @washumedicine's Dr. Daniel Marcus. intelligentimaging.ucsf.edu/events/xnat-2-…

A new study of 1,715 participants over 48 months in @OAC_Open by @UCSF_Ci2's Dr. Alexandra Gersing found thigh muscle loss was associated with greater tendon damage & decline in physical performance. oarsiopenjournal.com/article/S2665-…

Hospital archives offer anatomy & pathology diversity at scale, but their images are magnitude-only. @UCSFimaging #postdoc Nikhil Deveshwar shares how a generative model helps create synthetic multicoil MRI training data. ➡️ intelligentimaging.ucsf.edu/news/deep-gene…

Join us for our next SRG Pillar Meeting with @washumedicine's Dr. Daniel Marcus on XNAT 2. After over 2 decades supporting medical researchers, XNAT is adding agentic AI to research & extending its reach to enable responsible AI translation to the clinic. intelligentimaging.ucsf.edu/events/xnat-2-…