Bladder cancer (BCa) is one of the most common tumors in old males. It is usually classified into high-grade or low-grade according to the World Health Organization 2004 BCa grading criteria [1,2]. Tumor grade is a vital factor to guide therapeutic and follow-up decisions [3,4]. According to the National Comprehensive Cancer Network (NCCN) guidelines and European Association of Urology guidelines for BCa, high-grade tumors have higher recurrence and progression rates than low-grade tumors [5,6]. Moreover, once a high-grade tumor is confirmed, a second transurethral resection of the bladder tumor (TURBT) should be performed, and 1 to 3 years of adjuvant BCG instillation is recommended [7]. Thus, accurately predicting the BCa tumor grade preoperatively is crucial for treatment selection and clinical decision-making.

Currently, the histopathologic grading of BCa mainly depends on the findings of cystoscopic biopsy and TURBT [8], [9], [10], [11]. However, the tumor is heterogeneous, and the biopsy results are not always representative of the entire tumor and may require a repeated biopsy. Cystoscopic biopsy could provide tumor grade information pre-TURBT [12,13]. However, this approach is invasive and expensive. In addition, it is reported that 20% to 80% of the lesions were misdiagnosed because of variations in performing cystoscopic biopsy [14], [15], [16], [17]. Thus, setting up a noninvasive and accurate method to predict tumor grade preoperatively is urgently needed.

Published studies have shown that traditional magnetic resonance imaging (MRI), which includes T2-weighted imaging (T2WI), dynamic contrast-enhanced (DCE), and diffusion weighted imaging (DWI) sequences, plays an important role in diagnosing and predicting the grade of BCa, especially for mean apparent diffusion coefficient (mADC) mapping [18], [19], [20]. High-grade tumors have more active growth, density and irregular cell arrangement, which limit the intracellular and transmembrane mobility of water molecules. Since change in apparent diffusion coefficient (ADC) value is inversely proportional to cell fluidity, and it is negatively correlated with the tumor grade, and the mADC value may partially predict the histological grade of BCa. However, the mADC value overlaps between low-grade and high-grade tumors; it is the only predictive index, and a single mADC value is not powerful evidence and produces insufficient information to predict differentiation capabilities [21], [22], [23]. In addition, vesical imaging reporting and data system (VI-RADS) scoring, tumor size, and the number of tumors are also used to predict BCa grade in clinical practice [23]. The radiomics method, which is an advanced image-processing technique that extracts multiple quantitative features from images and is more objective and repeatable than traditional MRI [24,25], has been widely used in the diagnosis and preoperative grading of tumors of various systems in the body [26], [27], [28], [29], [30]. Several previous studies have demonstrated that the MRI radiomics models can be used to predict the histological grade of BCa before surgery [31], [32], [33]. However, as far as we know, no study has yet investigated the comparison between the potential of the biparametric (bp)-MRI-based radiomics model and that of the traditional MRI model (including mADC value, VI-RADS scoring, tumor size, and the number of tumors) in estimating the grade of BCa.

Therefore, the purpose of this study is 2-fold. First, we aimed to construct a radiomics model based on bp-MRI (ADC and T2WI) for the preoperative prediction of BCa grade. The second objective was to compare it with a traditional MRI model.