Talks on present and future CT technology kick off ISCT symposium

The last decade-plus has seen unprecedented development in all aspects of CT technology. Presenters will kick off the International Society for Computed Tomography symposium Tuesday morning May 18 looking at this evolution as part of the “Technology: Present and Future” session.

Dr. Geoff Rubin, a professor of radiology at Stanford University, will give an overview of this period, examining how sources of x-rays, the materials for detecting them, and the means for reconstructing data have changed over the last dozen years. Rubin will trace the development of CT to 1998 and the introduction of the first multidetector scanners, whose temporal and spatial resolution have since been dwarfed by devices with wide area detectors, multiple x-ray sources, high-pitch scanning, and iterative reconstruction methods that minimize dose while maximizing resolution.

With concerns mounting over patient radiation dose, the question arises: Can the CT community transition to low kV? Willi A. Kalender, Ph.D., a professor in the Institute of Medical Physics at the University of Erlangen in Germany will examine this question, looking into the optimal contrast-to-noise ratio. Kalender will present simulations and measurements regarding contrast, noise, and dose, focusing on spectral optimization for thoracic CT and cardiac CT.

Radiologists from Duke University and the University of Washington in Seattle will look at the argument for moving from filtered back projection to iterative reconstruction techniques. Dr. Rendon C. Nelson, a Duke professor of radiology, will describe the inner workings of filtered back projection (FBP) and two types of iterative reconstruction: adaptive statistical and model-based (MBIR). Nelson will compare dose and images obtained for varying exams.

Paul Kinahan, Ph.D., a UW professor of radiology and adjunct professor of bioengineering and electrical engineering, will focus on MBIR, the next generation of this type of algorithm for CT. He will present data indicating the effectiveness of MBIR at reducing noise levels compared with FBP, and its advantages in resolving soft tissue.

Dr. Patrik Rogalla, a professor in the University of Toronto medical imaging department, will describe a new twist on iterative reconstruction in the context of dual-source imaging. This adaptive iterative denoising reconstruction algorithm processes data from two imaging chains aligned at a 90º angle to each other. Rogalla will present the advantages of the technology, unique to Siemens’ high-end CTs, in speed and coverage, as well as spatial and temporal resolution. He will note the new clinical applications it supports, such as dual-energy, perfusion, and motion capture, and its potential for submilliSievert imaging of the heart.

Norbert J. Pelc, a professor of radiology and bioengineering at Stanford, will close the session with a description of progress achieved in the development of a radically different kind of CT, a multisource scanner built around inverse geometry. Multiple x-ray sources irradiate the target, in this case a phantom, creating a bow-tie pattern of x-rays projected onto a detector array, all rotating once per second.