Objective To investigate the potential of a two-component model combining T2 and diffusion weighted imaging(T2-DWI)in the diagnosis of prostate cancer.Methods Thirty-two prostate cancer patients and twenty prostatic hyperplasia patients underwent combined T2-DWI,with TE values set at 50 ms and 70 ms and b values at 50 s/mm2 and 800 s/mm2,respectively.This provided four measurements for each voxel.The signal fractions were calculated from the slow component(SFslow)extracted from the two-component model of T2-DWI,which was fitted with two free parameters.The results were compared with the traditional single exponential apparent diffusion coefficient(ADC)model.Results A significant difference was observed between SFslow(0.77±0.14 vs 0.35±0.10)and ADC[(0.91±0.24)μm2/ms vs(1.98±0.32)μm2/ms]in region of interest(ROI)between peripheral zone tumors and normal prostate tissue(P＜0.001).No significant difference was found between SFslow(0.75±0.13 vs 0.68±0.11)and ADC[(0.88±0.18)μm2/ms vs(0.97±0.13)μm2/ms]in ROI between non-peripheral zone tumors and prostatic hyperplasia.The area under the curve(AUC)of the receiver operating characteristic(ROC)for distinguishing tumors from prostate voxels was 0.962 for two-component SFslow and 0.940 for single exponential ADC models.The Spearman correlation coefficients between tumor SFslow and ADC with Gleason scores were 0.631 and-0.558,respectively.Conclusion SF estimation based on T2-DWI two-component model can effectively differentiate tumors from normal prostate tissue,showing potential in diagnosing prostate cancer.