A Novel Conversion Method from X-ray Image to MR Image Using Deep Network and Auto-Encoding Technology

Abstract

Several techniques have been utilized in current physical examinations, involving B-mode imaging, X-rays, magnetic resonance imaging (MRI), computed tomography (CT) scans, etc. Nevertheless, radiation exposure is delivered in variety of medical examinations such as MRI. Accordingly, yielding a high-quality medical image at the lowest possible radiation level becomes a realistic and challenging issue. This paper proposes a RVNet&PGAN method of simulation conversion based on deep network, which is capable of replacing the traditional methods by high-performance intelligent computing. Experimental results show that proposed algorithm performs better than other algorithms from multiple validity metrics such as MAE, CC and RMSE. From this perspective, the method has a salient effect on the conversion and reconstruction of medical images.