The basic idea of iterative image reconstruction is the calculation of image data that exactly corresponds to the obtained projection data. Applying the mathematical definition of an iterative algorithm for CT image reconstruction, the ideal IR procedure consists of a cycle of forward and backward projection steps with repeated transitions from projection (raw data) to image space and vice versa, iteratively improving the reconstructed image data. By simulating CT image acquisition using an accurate CT image model, step-forward projection generates synthetic projection data that is compared to measured projection data. The back-projection step propagates the correction determined from the difference between the simulated and measured projections to the image space (e.g. filtered back-projection) where it is applied as an update to the current estimate of the image data. In simple back projection (BP), the attenuation profile at each corner of the gantry is 'back-projected' across the image space. The attenuation value in the sinogram is divided by the number of image pixels along the projection direction from the x-ray source to the detector, and the average attenuation value thus obtained is assigned to these pixels; if the pixel values were summed along the projection direction, the original attenuation value in the sinogram would be returned. This is repeated for each corner of the gentry. The final back projection image is then the sum of all back projected attenuation profiles. Iterative reconstruction (IR) methods of MDCT images provide a reliable and alternative method to optimize the ratio between radiation dose and image quality. These methods consist of mathematical post-processing approaches that allow the correction of raw data by reducing image noise without changing the transverse spatial resolution.