An automatic method to detect unreliable BF values on CT images of lung cancers

Computed Tomography perfusion (CTp) is an imaging technique widely applied in the study of the efficacy of the new anti-angiogenetic therapies. The presence of blood vessels, bronchi, and artefacts jeopardizes the computation of perfusion indexes and the visual analysis of perfusion maps. In this work, we propose an automatic method for the detection of those anatomical structures and artefacts responsible of unreliable perfusion values to be excluded from the analysis.