Ingrid Framås Syversen^{1}, Mattijs Elschot^{2}, Tone Frost Bathen^{2}, and Pål Erik Goa^{3}

^{1}Kavli Institute for Systems Neuroscience, Norwegian University of Science and Technology, Trondheim, Norway, ^{2}Department of Circulation and Medical Imaging, Norwegian University of Science and Technology, Trondheim, Norway, ^{3}Department of Physics, Norwegian University of Science and Technology, Trondheim, Norway

T2 and diffusion-weighted imaging (DWI) are increasingly used in the detection and staging of prostate cancer, however, usually under the assumption that the T2 values and apparent diffusion coefficients (ADC) are independent of each other. Using hybrid multidimensional imaging, where images are acquired at two echo times and two b-values, we estimate volume fractions of slow and fast diffusion compartments in the prostate with a two-compartment model. The results suggest that the volume fractions can be used to discriminate between tumor and normal prostate tissue.

For each voxel, the 2×2 matrix of signal values S from the trace-weighted images was fitted to the following two-compartment model using Matlab (MathWorks, Natick, MA, USA):

$$\frac{S}{S_0}=V_{slow}\exp{\left(-\frac{TE}{T2_{slow}}\right)}\exp{\left(-b*ADC_{slow}\right)}+V_{fast}\exp{\left(-\frac{TE}{T2_{fast}}\right)}\exp{\left(-b*ADC_{fast}\right)},$$

where S

One tumor ROI and one normal tissue ROI were drawn for each patient. The median V

Median V

The hybrid multidimensional images can be acquired in a relatively short time, making it clinically feasible. The computational cost is also lower than using a three-compartment model. Furthermore, this model could potentially isolate the signal from subvoxel compartments and show underlying structural features of the tissue.

Although these initial results are encouraging, the model needs independent training, validation and testing in a larger number of patients and should also be correlated with histopathology. A possible source of bias in the results is that the ROIs were drawn by a basic scientist based on T2 and diffusion-weighted images and a radiologist description of the MRI exam. Furthermore, while all normal ROIs and nine tumor ROIs were in the peripheral zone, one tumor ROI was in the anterior fibromuscular stroma. In the future, the diagnostic properties of the model will be explored further, including correlation with Gleason scores and comparison with ADC.

1. Barentsz JO, Weinreb JC, Verma S, et al. Synopsis of the PI-RADS v2 Guidelines for Multiparametric Prostate Magnetic Resonance Imaging and Recommendations for Use. Eur Urol. 2016;69(1):41-49.

2. Wang S, Peng Y, Medved M, et al. Hybrid multidimensional T2 and diffusion-weighted MRI for prostate cancer detection. J Magn Reson Imaging. 2014;39(4):781-788.

3. Sadinski M, Karczmar G, Peng Y, et al. Pilot Study of the Use of Hybrid Multidimensional T2-Weighted Imaging-DWI for the Diagnosis of Prostate Cancer and Evaluation of Gleason Score. AJR Am J Roentgenol. 2016;207(3):592-598.

4. Ma X, Chen H, Peng Y. Hybrid T2-Weighted and Diffusion-Weighted Magnetic Resonance Imaging for Differentiating Prostate Cancer from Benign Prostatic Hyperplasia. In: 10th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics. Shanghai, China: Institute of Electrical and Electronics Engineers, 2017; 1-5.

5. Chatterjee A, Bourne RM, Wang S, et al. Diagnosis of Prostate Cancer with Noninvasive Estimation of Prostate Tissue Composition by Using Hybrid Multidimensional MR Imaging: A Feasibility Study. Radiology. 2018;287(3):864-873.

6. Karunamuni RA, Kuperman J, Seibert TM, et al. Relationship between kurtosis and bi-exponential characterization of high b-value diffusion-weighted imaging: application to prostate cancer. Acta Radiol. 2018;59(12):1523-1529.