Quantitative Assessment of Diffusion Spectrum Imaging in Breast Cancer

Overview

This project intends to prospectively collect patients with suspected breast malignant tumors by ultrasound or mammography. After routine MRI scanning, all patients underwent diffusion spectrum imaging (DSI) sequence scanning. The inclusion criteria were as follows: (1) breast cancer was confirmed by surgery or biopsy. (2) pathologically diagnosis of estrogen receptor (ER), progesterone receptor (PR), human epidermal growth factor 2 (HER-2), Ki-67, and lymphatic vessel invasion (LVI) status in breast cancer. (3) routine MRI and DSI scans were performed within one week before the pathologic examination. The exclusion criteria were as follows: (1) patients who had received treatment before DSI scanning; (2) patients who underwent breast tumor biopsy within two weeks before DSI image acquisition; (3) pathology results of breast masses were other diseases besides breast cancer. (4) post-processing of DSI data cannot be performed due to poor image quality, such as motion artifacts. Breast MRI data were collected on a 3T MR scanner (Magnetom skyra, Siemens Healthcare, Erlangen, Germany). All participants used standardized breast MRI scanning schemes, including T2 weighted imaging (T2WI), T1 weighted imaging (T1WI), diffusion-weighted imaging (DWI), DSI, and contrast dynamic enhancement (DCE). A total of 22 GSI quantitative parameters were derived from NeudiLab software that is based on the open-source platform DIPY (diffusion imaging in Python, http://nipy.org/dipy). The correlation between DSI quantitative parameters and pathological indexes (i.e., ER, PR, HER-2, Ki-67, and LVI) was evaluated by Spearman correlation analysis. The independent predictors of GSI quantitative parameters for different pathologic characteristics discrimination in breast cancer were determined by the logistic regression analysis. The predictive performance of DSI quantitative parameters for difference pathologic classifications was assessed by the receiver operating characteristic (ROC) curves and their respective area under the curves (AUCs).