Artificial intelligence in magnetic resonance imaging for predicting lymph node metastasis in rectal cancer patients: a meta-analysis.

Abstract

This meta-analysis aims to evaluate the diagnostic performance of magnetic resonance imaging (MRI)-based artificial intelligence (AI) in the preoperative detection of lymph node metastasis (LNM) in patients with rectal cancer and to compare it with the diagnostic performance of radiologists. A thorough literature search was conducted across PubMed, Embase, and Web of Science to identify relevant studies published up to September 2024. The selected studies focused on the diagnostic performance of MRI-based AI in detecting rectal cancer LNM. A bivariate random-effects model was employed to calculate pooled sensitivity and specificity, each reported with 95% confidence intervals (CIs). Study heterogeneity was assessed using the I² statistic. Furthermore, the modified quality assessment of diagnostic accuracy studies-2 (QUADAS-2) tool was applied to assess the methodological quality of the selected studies. Seventeen studies were included in this meta-analysis. The pooled sensitivity, specificity, and area under the curve (AUC) for MRI-based AI in detecting preoperative LNM in rectal cancer were 0.71 (95% CI: 0.66-0.74), 0.71 (95% CI: 0.67-0.75), and 0.77 (95% CI: 0.73-0.80), respectively. For radiologists, these values were 0.64 (95% CI: 0.49-0.77), 0.72 (95% CI: 0.62-0.80), and 0.74 (95% CI: 0.68-0.80). Both analyses showed no significant publication bias (p > 0.05). MRI-based AI demonstrates diagnostic performance similar to that of radiologists. The high heterogeneity among studies limits the strength of these findings, and further research with external validation datasets is necessary to confirm the results and assess their practical clinical value. Question How effective is MRI-based AI in detecting LNM in rectal cancer patients compared to traditional radiology methods? Findings The diagnostic performance of MRI-based AI is comparable to radiologists, with pooled sensitivity and specificity both at 0.71, indicating moderate accuracy. Clinical relevance Integrating MRI-based AI can enhance diagnostic efficiency in identifying LNM, especially in settings with limited access to skilled radiologists, but requires further validation.