1. Assessment of pathological grade and variants of bladder cancer with a continuous-time random-walk diffusion model
- Author
-
Wei Wang, Jingyun Wu, Qi Shen, Wei Li, Ke Xue, Yuxin Yang, and Jianxing Qiu
- Subjects
urinary bladder neoplasms ,pathology ,neoplasm grading ,tumor microenvironment ,diffusion magnetic resonance imaging ,Neoplasms. Tumors. Oncology. Including cancer and carcinogens ,RC254-282 - Abstract
PurposeTo evaluate the efficacy of high b-value diffusion-weighted imaging (DWI) with a continuous-time random-walk (CTRW) diffusion model in determining the pathological grade and variant histology (VH) of bladder cancer (BCa).MethodsA total of 81 patients (median age, 70 years; range, 35-92 years; 18 females; 66 high grades; 30 with VH) with pathologically confirmed bladder urothelial carcinoma were retrospectively enrolled and underwent bladder MRI on a 3.0T MRI scanner. Multi-b-value DWI was performed using 11 b-values. Three CTRW model parameters were obtained: an anomalous diffusion coefficient (D) and two parameters reflecting temporal (α) and spatial (β) diffusion heterogeneity. The apparent diffusion coefficient (ADC) was calculated using b0 and b800. D, α, β, and ADC were statistically compared between high- and low-grade BCa, and between pure urothelial cancer (pUC) and VH. Comparisons were made using the Mann–Whitney U test between different pathological states. Receiver operating characteristic curve analysis was used to assess performance in differentiating the pathological states of BCa.ResultsADC, D, and α were significantly lower in high-grade BCa compared to low-grade, and in VH compared to pUC (p < 0.001), while β showed no significant differences (p > 0.05). The combination of D and α yielded the best performance for determining BCa grade and VH (area under the curves = 0.913, 0.811), significantly outperforming ADC (area under the curves = 0.823, 0.761).ConclusionThe CTRW model effectively discriminated pathological grades and variants in BCa, highlighting its potential as a noninvasive diagnostic tool.
- Published
- 2024
- Full Text
- View/download PDF