Purpose: Methodological optimization of a 1H MR spectroscopic imaging sequence (1H-2D-CSI) and evaluation of its potential to diagnose prostate cancer as validated by histopathological maps.
Methods: The prostates of 18 patients were evaluated by 1H-MR-CSI (voxel dimension: 1 cm3) at 1.5 Tesla. This sequence was additionally combined with a frequency selective fat suppression.
Results: It was possible to distinguish prostate carcinoma from prostate hyperplasia spectroscopically by the ratio of citrate/(choline + creatine). Differentiation of high-grade prostatic intraepithelial neoplasia (PIN, high-grade) from prostate carcinoma was not unambiguously possible. Prediction of tumor differentiation was not possible by the ratio of citrate/(choline + creatine) by our maximum spatial resolution of 1 cm3.
Conclusion: 1H-2D-CSI is suitable for tumor detection. Tumor differentiation was not possible with the spatial resolution used.