Modified MR dispersion imaging in prostate dynamic contrast-enhanced MRI.

BACKGROUND: An estimation of an intravascular dispersion parameter was previously proposed to improve the overall accuracy and precision of the model parameters, but the high computation complexity can limit its practical usability in prostate dynamic contrast-enhanced MRI (DCE-MRI).

PURPOSE: To compare and evaluate the model fitting uncertainty and error in the model parameter estimation using different DCE-MRI analysis models and to evaluate the ability of the intravascular dispersion parameter to delineate between noncancerous and cancerous prostate tissue in the transition and peripheral zones.

STUDY TYPE: Retrospective.

POPULATION: Fifty-three patients who underwent radical prostatectomy.

FIELD STRENGTH/SEQUENCE: 3 T/3D RF-spoiled gradient echo sequence.

ASSESSMENT: The coefficient of variation was used to assess the model fitting uncertainty by adding random noise to the time-concentration curves, and the Akaike information criterion was used to assess the model fitting error. The parametric maps derived from four DCE-MRI analysis models were evaluated by evaluating the delineation between noncancerous tissue and prostate cancer or clinically significant prostate cancer.

STATISTICAL TESTS: The receiver operating curve analysis was performed to compare the ability to delineate between noncancerous and prostate cancer tissue in the transition and peripheral zones.

RESULTS: Both MR dispersion imaging (MRDI) and Weinmann analysis models had the maximum coefficient of variation in different tissue types, while the model fitting uncertainty of modified (m)MRDI was similar to the standard Toft model. In mMRDI, the model fitting error was minimum, and the delineation between noncancerous and clinically significant prostate cancer tissue was improved in both transition (area under the curve [AUC] = 0.92) and peripheral zones (AUC = 0.92), in comparison with MRDI (AUC = 0.89 and AUC = 0.85, respectively).

DATA CONCLUSION: The mMRDI showed promising results in detecting prostate cancer while maintaining a similar model fitting uncertainty.

LEVEL OF EVIDENCE: 3 Technical Efficacy: Stage 1 J. Magn. Reson. Imaging 2019.