Diagnostic performance of artificial intelligence in detection of renal cell carcinoma: a systematic review and meta-analysis
Abstract Objectives The detection of renal cell carcinoma (RCC) tumors in the earlier stages is of great importance for more effective treatment. Encouraged by the key role of imaging in the management of RCC, we conducted a systematic review and meta-analysis of the studies that made use of artific...
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
BMC
2025-01-01
|
| Series: | BMC Cancer |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s12885-025-13547-9 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Abstract Objectives The detection of renal cell carcinoma (RCC) tumors in the earlier stages is of great importance for more effective treatment. Encouraged by the key role of imaging in the management of RCC, we conducted a systematic review and meta-analysis of the studies that made use of artificial intelligence (AI) for the detection of RCC to quantitatively determine the performance of AI for distinguishing related renal lesions. Materials and methods PubMed, Scopus, CENTRAL, and Embase electronic databases were systematically searched in November 2024 to identify studies that applied AI for the detection or classification of RCC. We conducted a meta-analysis to evaluate the diagnostic performance of utilized algorithms. Moreover, meta-regression was conducted over suspected covariates to evaluate potential sources of inter-study heterogeneity. Publication bias and quality assessment were also done for the included studies. Results Sixty-four studies were included in this systematic review, of which 31 studies were selected for meta-analysis. The studies assessing algorithms’ performance on internal validation showed pooled sensitivity and specificity of 85% (95% confidence interval [CI], 82 to 87) and 76% (95% CI, 70 to 80), respectively. Moreover, externally validated Al algorithms had a pooled sensitivity and specificity of 80% (95% CI, 73 to 84) and 90% (95% CI, 84 to 93), respectively. Studies that performed internal validation for clinician performance had a pooled sensitivity of 79% (95% CI, 72 to 85) and specificity of 60% (95% CI, 49 to 70). Conclusion The findings of the present study validate the acceptable performance of AI algorithms when contrasted with medical professionals in the identification and categorization of RCC. Nevertheless, the presence of heterogeneity between studies and the absence of coherence in the results underscore the necessity for the cautious interpretation of these results and additional prospective studies. |
|---|---|
| ISSN: | 1471-2407 |