Ultrasound Radiomics for Differential Diagnosis of Renal Masses: A Systematic Review and Meta-Analysis

Abstract

Introduction Given the importance of early renal cancer diagnosis and the limitations of conventional imaging assessment, ultrasound radiomics has emerged as a promising non-invasive technique for differentiating benign from malignant renal masses. This meta-analysis is aimed to assess the ability of ultrasound radiomics in differentiating benign and malignant renal masses. Materials and Methods Literature was systematically retrieved from the inception of the databases to April 27, 2025. Meta-analysis was conducted using Review Manager, Meta-Disc, and STATA software to calculate pooled diagnostic accuracy metrics. Sensitivity analysis and subgroup analysis were performed to explore the sources of heterogeneity. Deeks’ funnel plot was adopted to evaluate publication bias, and the Fagan nomogram was applied to assess clinical utility. Results This study included 7 articles comprising a total of 1,864 cases (1,166 malignant and 698 benign). The pooled sensitivity, specificity, and AUC of ultrasound radiomics for differential diagnosis of benign and malignant renal masses were 0.83 (95% CI, 0.81-0.86), 0.79 (95% CI, 0.69-0.87), and 0.84(95% CI: 0.81-0.87). Significant heterogeneity was observed in pooled specificity among the 7 included studies. Subgroup analysis revealed that mass size and feature extraction software type contributed to the heterogeneity of specificity. Accordingly, 5 studies of simple small renal masses were further analyzed. The pooled sensitivity, specificity, and AUC of ultrasound radiomics for differential diagnosing of benign and malignant small renal masses were 0.84 (95% CI, 0.80-0.87), 0.72 (95% CI, 0.64-0.79), and 0.85(95% CI, 0.82-0.88). Fagan nomogram demonstrated a 30% increase in post-test probability for patients with positive pre-test results and a 15% decrease for those with negative pre-test results, indicating moderate clinical utility of radiomics. Conclusion Ultrasound radiomics has favorable value in the differential diagnosis of benign and malignant renal masses, and is expected to serve as a reliable auxiliary tool for clinical preoperative diagnosis.