Deep Convolutional approach for Low-Dose CT Image Noise Reduction

An essential objective in medical low-dose Computed Tomography (CT) imaging is how best to preserve the quality of the image. While, reducing the X-ray radiation dose is desired, in general, the image quality lowers by reducing the dose. Therefore, improving image quality is remarkably crucial for diagnostic purposes. A novel method to denoise low-dose CT images has been presented in this study. Different from the prevalent and traditional algorithms which utilize similar shared features of CT images in the spatial or transform domain, the deep learning approach is suggested for low-dose CT denoising. In this paper, a fully convolutional neural network architecture consisting of five parts, namely Feature extraction, Compressing, Mapping, Enlarging, and Assembling, are introduced to directly map the low-dose CT images onto the corresponding normal-dose CT images. The results of the proposed technique were compared with three state-of-the-art algorithms. To illustrate the superiority of our proposed technique, three performance measures, including root mean squared error, peak signal to noise ratio, and structural similarity index are presented.