Acquisition of Ktrans perfusion parameter maps from DCE-MRI in breast cancer using a deep learning approach

Background: Traditional methods for calculating perfusion parameter maps such as Ktrans require significant computational resources and time, leading to errors due to the variety of analysis models and the difficulty in estimating the arterial input function (AIF). Thus, it is difficult to apply Ktrans measurements to clinical diagnosis. The purpose of this study was to investigate whether deep learning (DL) techniques can synthesize Ktrans perfusion parameter maps from contrast-enhanced magnetic resonance (MR) images. Methods: A pix2pix-based conditional generative adversarial networks (cGAN) architecture was proposed to generate breast Ktrans perfusion maps. The peak signal-to-noise ratio (PSNR) and structural similarity index measure (SSIM) were used to evaluate the quality of the synthetic Ktrans maps. Two experienced radiologists were tasked with distinguishing between real and synthesized Ktrans maps. The Ktrans values of the tumor regions in the synthetic and real Ktrans maps were subjected to Pearson correlation analysis and Bland-Altman analysis. Results: The best performance was obtained when synthesizing Ktrans maps using the pix2pix model with spectral normalization (SN) and a local discriminator (LD) (PSNR of 15.167±0.125 and SSIM of 0.690±0.014 for synthesizing Ktrans maps from contrast-enhanced MR images). The Ktrans values of the tumor regions in the synthetic and real Ktrans maps showed a strong correlation (r=0.82), allowed for significant differentiation between benign and malignant tumors (P