A mode based de-convolution algorithm in CT cerebral tumor perfusion image

Abstract

With the advent of fast, multi-slice CT scanners, CT perfusion technique has significant implications in the diagnosis and treatment of cerebral tumor in the past 2 decades. The fundamental bases for dual chamber model which is selected to fit the tumor tissue are both the transport by blood flow of intravenously iodinated constrast material to tissue and the exchange by diffusion of these contrast between the intravascular space and the extravascular interstitial space. In this paper, a digital de-convolution algorithm is introduced and realized to compute the tissue residue function R(t), and the following parameters: cerebral blood flow (CBF), cerebral blood volume (CBV), mean transit time (MTT) and the tissue permeability coefficient (TPS). They can provide important reference before and after therapy. This algorithm is tested by CT cerebral tumors medical image data.