Chapter 27: Medical Imaging: CT, MRI, and Ultrasound

Learning Objectives

• Derive the Radon transform and the Fourier Slice Theorem, and explain why CT possesses an analytical inverse while RF imaging does not
• Formulate MRI as Fourier sampling with incomplete k-space coverage and apply compressed sensing (Lustig et al.) to accelerate acquisition
• Describe parallel imaging (SENSE, GRAPPA) and learned MRI reconstruction (E2E VarNet, fastMRI)
• Model ultrasound pulse-echo imaging as a near-field beamforming problem and connect delay-and-sum to the matched filter of Ch 13
• Identify which learned reconstruction architectures from medical imaging (MoDL, E2E-VarNet, learned primal-dual) transfer to RF imaging and what modifications the RF forward operator requires
• Articulate the ISAC paradigm as a capability unique to RF imaging with no current analogue in medical imaging