Radiomics of Treatment-naive Prostate Cancer Patients on Multiparametric MRI for Risk Stratification and Treatment Outcomes Predictions

Prostate cancers (PCA) are a heterogeneous group which include indolent tumors that has no clinical significance to very aggressive cancer that could result in morbidities and mortality. Thus, an accurate risk stratification at the time of PCA diagnosis is crucial. The histological examination of PCA biopsy specimens could not accurately predict the final tumor aggressiveness shown on radical prostatectomy specimens because of heterogeneous distributions of the most malignant tumor cells. Prostate multiparametric magnetic resonance imaging (mpMRI) has been generally accepted to be the best imaging modality for detecting and localizing prostate cancers themselves. Furthermore, the rapid development of radiomics provide comprehensive quantitative information of all tumor data which could be used for risk stratification and prognosis prediction. Thus, this study plans to enroll 200 eligible patients who undergo prostate mpMRI first, followed by radical prostatectomy for prostate cancers. We use radiomics extracted from prostate mpMRI for risk stratification patients of histological aggressiveness as well as to predict very early recurrence of PCA patients within 6 months after radical prostatectomy.

Sex

Male

Minimum age

20 Years

Inclusion Criteria:

1. Aged over 20 years old.
2. Suspected or confirmed prostate cancer.
3. Undergoing prostate mpMRI before clinical treatment.
4. Normal renal function(i.e.: estimated GFR ≧60).
5. No allergy history to gadolinium based contrast agent.
6. Agree to participate this study and sign informed consent.

Exclusion Criteria:

1. mpMRI photography not completed.
2. mpMRI images are damaged or poor in quality and cannot be interpreted.
3. Without pathological examination confirmed prostate cancer.
4. Patient withdraw informed consent.