Purpose The purposes of this study were to evaluate the feasibility of simultaneous ¹⁸F choline positron emission tomography (PET) and magnetic resonance imaging (MRI) of the prostate and to present the first clinical results of the method. Materials and Methods From March 2012 to October 2012, a total of 15 consecutive patients were examined with simultaneous ¹⁸F choline PET/MRI. At the time of the examination, 8 patients had histologically proven prostate cancer, 2 patients had repeated prostate biopsies with negative results, and 5 patients had suspected prostate cancer with an elevated or rising prostate specific antigene level but did not have a prostate biopsy. Sequence protocol comprised T2-weighted high-resolution images and diffusion-weighted images of the prostate in addition to PET imaging. Image quality was assessed by 2 radiologists, and the PET images were evaluated qualitatively and quantitatively. Results Simultaneous PET/MRI of the prostate was accomplished successfully in all patients. The method proved to be robust without technical failure, and the image quality was rated to be diagnostic in all examinations except in 1 diffusion-weighted imaging (DWI) data set that was judged to be nondiagnostic because of susceptibility artifacts. High-resolution T2-weighted images allowed exact correlation of elevated focal or diffuse choline uptake to suspicious T2-weighted lesions of the prostate. A high accordance was found between PET and DWI. However, PET-positive lesions were found in 3 patients wherein DWI did not indicate tumor in suspicious T2-weighted lesions. Conclusions Simultaneous positron emission tomography/magnetic resonance imaging of the prostate has the advantage of combining high-resolution prostate images, functional studies, and metabolic/molecular imaging. The PET component adds diagnostic confidence to the MRI-based parameters in identifying and localizing tumor in the prostate.

Background: Diffusion-weighted imaging (DWI) is routinely used in magnetic resonance imaging (MRI) of prostate cancer. However, the routine use of b values higher than 1000 s/mm² is not clear up to present. Moreover, the complex diffusion behavior of malignant and benign prostate tissues hampers precise predictions of contrast in DWI images and apparent diffusion coefficient (ADC) maps. Purpose: To quantitatively analyze DWI with different b values in prostate cancer and to identify b values best suitable for cancer detection. Material and Methods: Forty-one patients with histologically proven prostate cancer were examined with high resolution T2-weighted imaging and DWI at 3 Tesla. Five different b values (0, 800, 1000, 1500, 2000 s/mm²) were applied. ADC values of tumors and reference areas were measured on ADC maps derived from different pairs of b values. Furthermore, signal intensities of DW images of tumors and reference areas were measured. For analysis, contrast ratios of ADC values and signal intensities of DW images were calculated and compared. Results: No significant differences were found between contrast ratios measured on ADC maps of all analyzed b value pairs ( P = 0.43). Contrast ratios calculated from signal intensities of DW images were highest at b values of 1500 and 2000 s/mm² and differed significantly from contrast ratios at b values of 800 and 1000 s/mm² ( P < 0.01). Conclusion: Whereas contrast in ADC maps does not significantly change with different b values, contrast ratios of DW images are significantly higher at b-values of 1500 and 2000 s/mm² in comparison to b values of 800 and 1000 s/mm². Therefore, diagnostic performance of DWI in prostate cancer might be increased by application of b values higher than 1000 s/mm².

To characterize intermediate and high-risk prostate carcinomas with measurements of standardized uptake values (SUVs) and apparent diffusion coefficient (ADC) values by means of simultaneous [¹⁸F] choline PET/MRI. 35 patients with primary prostate cancer underwent simultaneous [¹⁸F] choline PET/MRI. From these, 21 patients with an intermediate and high risk constellation who were not under ongoing hormonal therapy were included. Altogether 32 tumor lesions with a focal uptake of [¹⁸F] choline could be identified. Average ADC values (ADC_aver) minimum ADC values (ADC_min) as well as maximum and mean SUVs (SUV_max, SUV_mean) of tumor lesions were assessed with volume-of-interest (VOI) and Region-of-interest (ROI) measurements. As a reference, also ADC_aver, ADC_min and SUV_max and SUV_mean of non-tumorous prostate tissue were measured. Statistical analysis comprised calculation of descriptive parameters and calculation of Pearson’s product moment correlations between ADC values and SUVs of tumor lesions. Mean ADC_aver and ADC_min of tumor lesions were 0.94±0.22×10⁻³ mm²/s and 0.65±0.21×10⁻³ mm²/s, respectively. Mean SUV_max and SUV_mean of tumor lesions were 6.3±2.3 and 2.6±0.8, respectively. These values were in each case significantly different from the reference values (p<0.001). There was no significant correlation between the measured SUVs and ADC values (SUV_max vs. ADC_aver: R = −0.24, p = 0.179; SUV_max vs. ADC_min: R = −0.03, p = 0.877; SUV_mean vs. ADC_aver: R = −0.27, p = 0.136; SUV_mean vs. ADC_min: R = −0.08, p = 0.679). Both SUVs and ADC values differ significantly between tumor lesions and healthy tissue. However, there is no significant correlation between these two parameters. This might be explained by the fact that SUVs and ADC values characterize different parts of tumor biology.

Short tau inversion recovery (STIR) sequences are frequently used in magnetic resonance imaging (MRI) of the spine. However, STIR sequences require a significant amount of scanning time. The purpose of the present study was to generate virtual STIR (vSTIR) images from non-contrast, non-fat-suppressed T1- and T2-weighted images using a conditional generative adversarial network (cGAN). The training dataset comprised 612 studies from 514 patients, and the validation dataset comprised 141 studies from 133 patients. For validation, 100 original STIR and respective vSTIR series were presented to six senior radiologists (blinded for the STIR type) in independent A/B-testing sessions. Additionally, for 141 real or vSTIR sequences, the testers were required to produce a structured report of 15 different findings. In the A/B-test, most testers could not reliably identify the real STIR (mean error of tester 1–6: 41%; 44%; 58%; 48%; 39%; 45%). In the evaluation of the structured reports, vSTIR was equivalent to real STIR in 13 of 15 categories. In the category of the number of STIR hyperintense vertebral bodies (p = 0.08) and in the diagnosis of bone metastases (p = 0.055), the vSTIR was only slightly insignificantly equivalent. By virtually generating STIR images of diagnostic quality from T1- and T2-weighted images using a cGAN, one can shorten examination times and increase throughput.