Predicting the aggressiveness of peripheral zone prostate cancer using a fractional order calculus diffusion model. Issue 143 (October 2021)

• Record Type: Journal Article
• Authors: Li, Zhihua; Dan, Guangyu; Tammana, Vikram; Johnson, Scott; Zhong, Zheng; Rabiee, Behnam; Zhou, Xiaohong Joe; L. Xie, Karen
• Is Part Of: European journal of radiology. Issue 143(2021)
• Publication Date: 2021-10
• Subjects: Medical radiology -- Periodicals; Radiology -- Periodicals; Radiologie médicale -- Périodiques; Medical radiology; Periodicals; 616.075705
• DOI: 10.1016/j.ejrad.2021.109913
• Languages: English
• ISSNs: 0720-048X

Abstract

Highlights: FROC display higher performance than ADC in differentiating PCa grade in low (GS 6) vs. intermediate (GS 7) risk PCa and clinically insignificant (GS 6) vs. significant (GS ≥ 7) PCa. Combination of FROC parameters provide higher performance than individual FROC parameters and ADC. FROC D, which measures mean diffusion, is inversely correlated with PCa grade. The value of FROC β, which inversely correlates with diffusion heterogeneity, is lowest in GS 7 PCa. Abstract: Purpose: To evaluate the performance of parameters D, β, μ from the Fractional Order Calculus (FROC) model at differentiating peripheral zone (PZ) prostate cancer (PCa) Material and Methods: 75 patients who underwent targeted MRI-guided TRUS prostate biopsy within 6 months of MRI were reviewed retrospectively. Regions of interest (ROI) were placed on suspicious lesions on MRI scans. ROIs were then correlated to pathological results based on core biopsy location. The final tumor count is a total: 23 of GS 6 (3 + 3), 36 of GS 7 (3 + 4), 18 of GS 7 (4 + 3), and 19 of GS ≥ 8. Diffusion-weighted imaging (DWI) scans were fitted into the FROC and monoexponential model to calculate ADC and FROC parameters: anomalous diffusion coefficient D, intravoxel diffusion heterogeneity β, and spatial parameter μ. The performance of FROC parameters and ADC at differentiating PCa grade was evaluated with receiver operating characteristic (ROC) analysis. Results: In differentiating low (GS 6) vs. intermediate (GS 7) risk PZ PCa, combination of (D, β) provides the best performance with AUC of 0.829 with significance of p = 0.018 when compared to ADC (AUC of 0.655). In differentiating clinically significant (GS 6) vs. clinically significant (GS ≥ 7) PCa, combination of (D, β, μ) provides highest AUC of 0.802 when compared to ADC (AUC of 0.671) with significance of p = 0.038. Stratification of intermediate (GS 7) and high (GS ≥ 8) risk PCa with FROC did not reach a significant difference when compared to ADC. Conclusion: Combination of FROC parameters shows greater performance than ADC at differentiating low vs. intermediate risk and clinically insignificant vs. significant prostate cancers in peripheral zone lesions. The FROC diffusion model holds promise as a quantitative imaging technique for non-invasive evaluation of PZ PCa.