Purpose: To evaluate the feasibility of using mpMRI image features predicted by AI algorithms in the prediction of clinically significant prostate cancer (csPCa). Materials and methods: This study analyzed patients who underwent prostate mpMRI and radical prostatectomy (RP) at the Affiliated Hospital of Jiaxing University between November 2017 and December 2022. The clinical data collected included age, serum prostate-specific antigen (PSA), and biopsy pathology. The reference standard was the prostatectomy pathology, and a Gleason Score (GS) of 3 + 3 = 6 was considered non-clinically significant prostate cancer (non-csPCa), while a GS ≥ 3 + 4 was considered csPCa. A pre-trained AI algorithm was used to extract the lesion on mpMRI, and the image features of the lesion and the prostate gland were analyzed. Two logistic regression models were developed to predict csPCa: an MR model and a combined model. The MR model used age, PSA, PSA density (PSAD), and the AI-predicted MR image features as predictor variables. The combined model used biopsy pathology and the aforementioned variables as predictor variables. The model’s effectiveness was evaluated by comparing it to biopsy pathology using the area under the curve (AUC) of receiver operation characteristic (ROC) analysis. Results: A total of 315 eligible patients were enrolled with an average age of 70.8 ± 5.9. Based on RP pathology, 18 had non-csPCa, and 297 had csPCa. PSA, PSAD, biopsy pathology, and ADC value of the prostate outside the lesion (ADC_prostate) varied significantly across different ISUP grade groups of RP pathology (P < 0.001). Other clinical variables and image features did not vary significantly across different ISUP grade groups (P > 0.05). The MR model included PSAD, the ratio of ADC value between the lesion and the prostate outside the lesion (ADC_{lesion/prostate}), the signal intensity ratio of DWI between the lesion and the prostate outside the lesion (DWI_{lesion/prostate}), and the ratio of DWI_{lesion/prostate} to ADC_{lesion/prostate}. The combined model included biopsy pathology, ADC_{lesion/prostate}, mean signal intensity of the lesion on DWI (DWI_lesion), DWI signal intensity of the prostate outside the lesion (DWI_prostate), and signal intensity ratio of DWI between the lesion and the prostate outside the lesion (DWI_{lesion/prostate}). The AUC of the MR model (0.830, 95% CI 0.743, 0.916) was not significantly different from that of biopsy pathology (0.820, 95% CI 0.728, 0.912, P = 0.884). The AUC of the combined model (0.915, 95% CI 0.849, 0.980) was higher than that of the biopsy pathology (P = 0.042) and MR model (P = 0.031). Conclusion: The aggressiveness of prostate cancer can be effectively predicted using AI-extracted image features from mpMRI images, similar to biopsy pathology. The prediction accuracy was improved by combining the AI-extracted mpMRI image features with biopsy pathology, surpassing the performance of biopsy pathology alone.

The use of deep learning radiomics to predict whether advanced prostate cancer (PCa) will progress within two years after treatment has been validated, yet there remains a lack of research on estimating time to progression. Patients were enrolled from October 2017 to March 2024. One hundred and eighty-two patients with advanced PCa diagnosed through ultrasound-guided systematic prostate biopsy were enrolled. A deep learning-based radiomics model for predicting progression was firstly developed using pretreatment MR apparent diffusion coefficient (ADC) maps, and the performance of manual (ROIref) versus AI-derived (ROIai) tumor segmentations was compared. Then, survival analysis was performed to compare ROIref-based and ROIai-based radiomics-predicted probabilities in the risk stratification. The area under the receiver operating characteristics curve (AUC) was used to estimate the model efficacy. The model achieved high AUC values for progression prediction in test sets (ROIref: 0.840, ROIai: 0.852). No significant difference was observed between ROIai-based and ROIref-based approaches (ΔAUC = 0.012, p = 0.870) in the test set. Both ROIref-predicted and ROIai-predicted probabilities independently predicted progression in multivariate Cox proportional hazard regression models (p < 0.001) and stratified patients into distinct survival groups (log-rank p < 0.001). Decision curve analysis confirmed equivalent clinical utility across thresholds (0.1–0.6), with net benefit exceeding the “treat all” and “treat none” strategies. In conclusion, deep learning-based radiomics models could effectively predict advanced PCa progression, with AI-derived tumor annotations performing equally to manual expert ones.