UC Berkeley

Magnetic Particle Imaging (MPI) is an emerging imaging modality with potential clinical applications in rapid angiography, cell therapy tracking, cancer imaging, and inflammation imaging. While still in its infancy, MPI already has excellent contrast, safety, depth penetration, and sensitivity without the serious health risks posed by current modalities such as ionizing radiation and iodinated contrast agents.

With any tracer imaging modality, high sensitivity is necessary to improve safety and enable new clinical applications. The first Magnetic Computed Tomography (MCT) system and experimental images were created in this work with the goal of improving sensitivity in MPI. MCT uses projection reconstruction algorithms similar to those in X-ray computed tomography to reconstruct 3D images from a projection MPI sytem. Analytical derivation and experimental evidence demonstrate that MCT has an order-of-magnitude higher SNR than previous 3D MPI methods for the same scan time as well as a 40% resolution improvement. The MCT experimental system acquisition speed was improved from initial work to acquire images twenty-fold faster at less than two minutes per 3D image.

Critical to any medical imaging technology is the reliability and accuracy of image reconstruction. Unfortunately, prior approaches to x-space MPI reconstruction suffer from image artifacts such as banding and haze. In this work, a priori knowledge of image continuity and non-negativity are introduced into a new optimization formulation to reduce these non-physical artifacts in 2D and 3D reconstructions.