Compressive computed tomography image reconstruction by using the analysis of the internal structure of an object

Computed Tomography (CT) is a non-invasive and non-intrusive technique that allows classification and detection of the internal structure of an object. However, the high doses of radiation generated by CT scanners are excessive, and it may represent a risk to the patient's health or even damage to the object of study. To reduce this damage is necessary to decrease the doses of radiation, i.e., lowering the number of view angles at which projections are taken. However, the reduction of measurements leads to an inverse ill-posed inverse problem. Coded aperture X-ray tomography is an approach that allows to overcome these limitations. This approach is based on the Compressive Sensing (CS) theory, which emerged as a new sampling technique requiring fewer projections than those specified by the Nyquist criterion. However, CS method in CT does not exploit the internal structure of the object. In this paper, we propose a strategy of CS in CT using adaptive coded aperture to obtain better reconstruction of CT images. Coded apertures are adapted using an initial reconstruction of the object of interest that is obtained from a previous shot. The results indicate that by using just 18% of the samples, it is possible to obtain up to 2 dB improvement in terms of PSNR (Peak-signal-to-noise-ratio) in reconstructed images compared to the traditional method.