Keywords:
Abdomen, Computer applications, Liver, CT-Quantitative, CT, Computer Applications-General, Imaging sequences, Technical aspects, Tissue characterisation, Cancer
Authors:
M. Mottola, A. Bevilacqua; Bologna/IT
DOI:
10.26044/ecr2019/C-0350
Aims and objectives
In oncology,
the researchers are aiming at consolidating the development of target treatments employing antiangiogenic drugs that act on specific biological pathways in order to limit or prevent the arrival of oxygen and nutrients at tumour or metastatic sites [1].
In fact,
these drugs causing a cytostatic rather than a cytotoxic effect fight against the vascular reshaping,
occurring since the early phases of tumour growth [2],
thus making the traditional morphological techniques unsuitable to assess such therapies response [3].
Computed Tomography perfusion (CTp) is a very promising functional imaging technique able to depict the microvascular characteristics of tissue by following the passage of a Contrast Agent (CA) within the volume of interest through repeated CT scans.
In particular,
CTp provides Time Concentration Curves (TCCs) which are further used to measure the CA enhancement in Hounsfield Unit (HU) [4].
These signals are exploited in order to compute several perfusion parameters,
the most relevant of which is the Blood Flow (BF) [5].
However,
at present CTp is not standardized in the clinical routine due to different analytical methods yielding non-comparable results [6].
As regards the computation of BF,
one of the simplest and adopted methods is the Maximum Slope [7].
In particular,
the Maximum Slope (MS) measures the BF referred to the first-pass kinetics of CA,
when the venous outflow has yet to occur.
However,
CA delays potentially sap the validity of the no-venous outflow assumption,
thus increasing the level of uncertainty and approximation of the BF achieved via MS.
In this work,
we exploit tissue TCCs to compute a new perfusion feature,
the equilibrium BF (eBF) that is a univocal descriptor for each-voxel based TCC,
corresponding to the CA balance of at the system's inlet and outlet.
eBF and MS-based BF performance are compared through the Coefficient of Variation (CV) and voxel-based correlations are assessed through the squared Pearson’s index (R2).
As a matter of fact,
eBF,
which allows for outflow measurements,
is a normalizing feature since all TCCs at the time instant of their maximum value offer a representation of the same state of the system.
This may enrich clinical assessments and improve reliability of perfusion values,
urging CTp standardization.