Radiomics for Detection of Prostate Cancer in PI-RADS 3 Lesions: A Systematic Review and Meta-analysis.

RATIONALE AND

OBJECTIVES

The noninvasive diagnosis of prostate cancer (PCa) in Prostate Imaging-Reporting and Data System (PI-RADS) 3 lesions remains challenging. The effectiveness of radiomics for detecting PCa in PI-RADS 3 lesions has been explored in some studies, but systematic evidence is still lacking to date. Therefore, this study was conducted to synthesize the effectiveness of radiomics for detecting PCa in PI-RADS 3 lesions.

MATERIALS AND

METHODS

A systematic search was conducted on Embase, PubMed, Cochrane Library, and Web of Science until September 15, 2024 for studies on the detection performance of radiomics for all PCa and clinically significant PCa (csPCa) in PI-RADS 3 lesions. Quality assessment was performed using the radiomics quality score (RQS).

RESULTS

Twelve studies were included, encompassing 2367 PI-RADS 3 lesions in 2340 patients. Meta-analyses revealed that in the validation set, the pooled sensitivity, specificity, and area under the summary receiver operating characteristic curve (AUC) of radiomics were 0.76 (95%CI 0.71-0.80), 0.79 (95%CI 0.72-0.85), and 0.78 (95%CI 0.32-0.96), respectively, for detecting all PCa. In the validation set, the pooled sensitivity, specificity, and AUC of radiomics were 0.76 (95%CI 0.71-0.80), 0.82 (95%CI 0.74-0.88), and 0.77 (95%CI 0.49-0.92), respectively, for detecting csPCa.

CONCLUSION

Radiomics possesses high accuracy in the detection of PCa in PI-RADS 3 lesions. However, both the number and quality of the original studies included were limited, so it is necessary to conduct more prospective, large-sample, multicenter studies in the future.