Erratum to: Principles and applications of multienergy CT: Report of AAPM Task Group 291 (Medical Physics, (2020), 47, 7, 10.1002/mp.14157)

In AAPM Report 291, “Principles and applications of multienergy CT: Report of AAPM Task Group 291,”¹ scientists from Canon Medical Systems Corporation, GE Healthcare Technologies, Philips Healthcare, and Siemens Healthcare GmbH participated as authors and contributed sections to describe, in general, the technology used in their company’s dual- or multi-energy CT systems. Specific scanner models were sometimes mentioned to provide examples of the overall approach. The general statements made about the approach may not, however, apply to the specific scanner model used as an example. The intermingling of the general and specific statements is regrettable, as it is ready source of potential confusion. Readers should view each section as a description of an overall dual- or multi-energy CT approach, even if a specific scanner model is mentioned as an example. Readers should refer to manufacturer specifications and the published literature to assess the specifications and performance of a specific CT system. A numerical error exists on page e895 in the second paragraph of the limitations section for kV switching. The statement “… so the system software currently prevents the use of gantry rotation times of less than approximately 500 ms, which may limit the quality of cardiac examinations,” implies that a minimum 500 ms rotation time applies to the scanner used as an example in the kV switching setting (GE Discovery CT750 HD). However, this value (500 msec) does not reflect the minimum rotation time on the GE Discovery CT750 HD. Rather, from the earliest publications on the use of this specific CT scanner model for dual-energy cardiac imaging, the minimum rotation time of 350 ms was used.^2,3 The second and third paragraphs in this same section on the limitations of kV switching reiterate the physics principles discussed on page e894, at the top of the second. Namely, as first demonstrated in 1979 by Kelcz et al., decreased spectral separation causes increased noise in decomposed material-specific images.⁴ In 2009, Primak et al. demonstrated that tube-specific filtration can increase spectral separation and decrease image noise in material decomposed images using a commercial dual-source CT system.⁵ While spectral separation is indeed an important determinant of the underlying noise properties of the acquired data, noise reduction methods can be applied such that the material decomposition calculations performed using kV switching (or other non-dual-source approaches) can also yield accurate quantification of iodine concentration.⁶.