Double Transformer Super-Resolution for Breast Cancer ADC Images

Diffusion-weighted imaging (DWI) has been extensively explored in guiding the clinic management of patients with breast cancer. However, due to the limited resolution, accurately characterizing tumors using DWI and the corresponding apparent diffusion coefficient (ADC) is still a challenging problem. In this paper, we aim to address the issue of super-resolution (SR) of ADC images and evaluate the clinical utility of SR-ADC images through radiomics analysis. To this end, we propose a novel double transformer-based network (DTformer) to enhance the resolution of ADC images. More specifically, we propose a symmetric U-shaped encoder-decoder network with two different types of transformer blocks, named as UTNet, to extract deep features for super-resolution. The basic backbone of UTNet is composed of a locally-enhanced Swin transformer block (LeSwin-T) and a convolutional transformer block (Conv-T), which are responsible for capturing long-range dependencies and local spatial information, respectively. Additionally, we introduce a residual upsampling network (RUpNet) to expand image resolution by leveraging initial residual information from the original low-resolution (LR) images. Extensive experiments show that DTformer achieves superior SR performance. Moreover, radiomics analysis reveals that improving the resolution of ADC images is beneficial for tumor characteristic prediction, such as histological grade and human epidermal growth factor receptor 2 (HER2) status.

I. Introduction

Breast cancer is the most prevalent disease with a high mortality rate that has affected women's health worldwide in the past few decades. Fortunately, scientific evidence has repeatedly supported that early detection and treatment can significantly improve the survival rate of patients. Dynamic contrast-enhancement magnetic resonance imaging (DCE-MRI) has emerged as an evolving imaging technique for cancer screening and detection, which exhibits significantly higher sensitivity in unclear and ambiguous breast lesions [1]. DCE-MRI can not only portray the morphological characteristics of breast lesions, but also reveal the physiological changes and dynamic enhancement properties of contrast agents in cancer. To date, DCE-MRI has been widely used and exhibited promising practicality in the identification and characterization of early cancers depicted in fatty and dense breasts [2], [3]. In clinical settings, the radiomics spatiotemporal association features of DCE-MRI have exhibited high utility for breast cancer diagnosis and treatment [4], such as identification of the histological grades [5] and ki-67 expression status [6], [7], the response and prognosis to chemotherapy [8], [9], and molecular subtype [10], [11], [12].