The effects of baseline length in Computed Tomography perfusion of liver

Abstract Objective: Computed Tomography perfusion (CTp) of liver is very attractive for predictive and prognostic purposes, but motion artefacts and radiation dose connected to duration of examinations jeopardize the reproducibility of perfusion values, thwarting CTp daily application in clinics. The goal is showing to what extent these issues can be faced by shortening the CTp unenhanced stage (i.e., the baseline). Methods: 59 patients with colorectal cancer underwent undelayed hepatic CTp examinations. For each patient, fifteen virtual examinations E τ simulating different scan delays τ ∈ [1..15] s were achieved from the undelayed original sequence E 0 . Absolute (AD), percentage (PD) and compound differences ( CD τ ) were computed between E 0  and each E τ for baseline and perfusion values and measured in HU and arbitrary units (a.u.), respectively. Patients were grouped and counted based on the differences achieved. Results: Maximum perfusion CD τ 10 a.u. and baseline CD τ 7 HU were achieved. For τ ≤ 10 s, maximum perfusion CD τ ∈ [5,6) a.u. was found in one patient only as well as maximum baseline CD τ ∈ [2,3) HU. Blood flow (BF), hepatic perfusion index and arterial BF showed the lowest CD τ , while portal BF and total BF the highest ones. PD is practically always higher than AD. Conclusion: The approach presented allows clinicians to design the shortest CTp acquisition protocol, selecting the highest delay compliant with the required accuracy for the chosen perfusion parameters, to limit patient’s motion and improve image quality. Significance: A short CTp protocol allows strengthening the reliability of perfusion values, and correctness of clinical outcomes, advancing CTp introduction in the standard clinical practice.