Keywords:
Liver, Artificial Intelligence, Oncology, CT, CT-Quantitative, Computer Applications-Detection, diagnosis, Computer Applications-General, Imaging sequences, Cancer, Metastases, Tissue characterisation
Authors:
M. Mottola1, A. Bevilacqua1, V. Vilgrain2; 1Bologna/IT, 2Clichy/FR
DOI:
10.26044/ecr2019/C-0522
Aims and objectives
Colorectal cancer yields significant morbidity and mortality and it is often associated to metastatic diseases,
commonly localized in liver parenchyma.
As a general tendency,
half of all patients with colorectal cancer will develop liver metastases [1].
Malignant changes of tissues are known to cause vascular reshaping and neoangiogenic processes occurring to ensure the transport of nutrients and oxygen and favour tumour growth [2].
Therefore,
the ability to early detect regional alterations in organs’ hemodynamics is one of the major challenges of standard medical imaging techniques,
where perfusion imaging can show all its potentiality [3].
Among available tools for perfusion measurements,
the Dynamic Contrast Enhanced Computed Tomography (DCE-CT or CTp) is a very promising functional technology,
also thanks to its availability in most of hospitals,
due to the reduced costs,
and patients’ acceptability [4].
CTp provides a functional characterization of tissues through repeated CT scans performed before,
during,
and after the intravenous injection of a Contrast Agent (CA),
in order to follow its temporal evolution within the tissue Region of Interest (ROI) [5] through the recovery of the Time-Concentration Curves (TCCs).
Analytical methods for the computation of perfusion parameters exploit the TCCs to derive image-based biomarkers,
thus allowing assessing tumour changes and treatments response [4].
However,
motions artefacts and image noise can seriously compromise the quality of the TCCs extracted and jeopardize perfusion measurements [6].
In fact,
despite its recognised usefulness,
CTp is not yet standardized in the clinical routine,
due to several perfusion methods yielding non-comparable results [7].
This work aims at providing a practical proof of CTp potentiality for carrying out quantitative measurements of tissue perfusion able to yield potential biomarkers of the development of liver metastases in patients with colorectal cancer.
Six perfusion parameters were computed and 24 statistical features were extracted and coupled,
thus attaining 276 possible combinations.
Results show the predictivity of a Hepatic Perfusion Index (HPI)-based feature to linearly separate patients who will develop metastases with maximum specificity.
Ultimately,
this is expected to strong push the interest for CTp standardization.