Volumetric tumor segmentation according to diffusion-weighted MRI data in predicting and evaluating the response to chemotherapy in bladder cancer

Abstract

The purpose of the study was to develop and evaluate the technique for volumetric tumor segmentation based on diffusion-weighted magnetic resonance imaging (DW-MRI) in predicting and assessing the response to chemotherapy in patients with bladder cancer (BC). Material and Methods. We examined 26 patients with morphologically verified transitional cell carcinoma of the bladder. The group was characterized by the presence of one or several tumors with a size of 17 to 46 mm. Before planning chemotherapy according to the M-VAC scheme, a study and post-processing of DW-MRI with volumetric segmentation of lesions, assessment of the volume and apparent diffusion coefficient (ADC) in the entire tumor mass were performed. According to the ADC data, shape of the tumor and its relation to the bladder wall, the coefficient (C) for predicting the response to chemotherapy was calculated. Results. In the cases with a coefficient value below 0.51, a high risk of treatment failure was predicted, at C≥0.74, a positive effect of treatment was predicted. With a value of 0.51≤C<0.74, the prognosis was uncertain, stabilization was more likely. The sensitivity and specificity of the method were 92.3 % and 92.4 %, respectively. Conclusion. The method of volumetric segmentation makes it possible to predict and evaluate bladder cancer response to chemotherapy with a sufficiently high accuracy. The advantages of the method are the possibility of assessing the tumor regardless of the degree of filling of the bladder, with non-contrast MRI, and with large lesions.