Department of Ultrasound in Medicine, Ningbo Medical Center Lihuili Hospital, No.1111, Jiangnan Road, Yinzhou District, Ningbo 315048, Zhejiang Province, China , yanyun_1975@163.com
Abstract: (9 Views)
Background:To evaluate the diagnostic efficacy of Magnetic resonance imaging (MRI) radiomics in predicting the overexpression of Human Epidermal Growth Factor Receptor 2(HER2) in breast cancer via meta-analysis, so as to investigate the correlation between MRI radiomics and HER2 overexpression in breast cancer. Materials and Methods: PubMed, Embase, and Web of Science were searched for studies on radiomics and HER2 overexpression in breast cancer (inception–September 2025). Two reviewers independently screened literature, extracted data, and assessed quality (RevMan 5.3). Pooled sensitivity (SEN), specificity (SPE), and AUC were calculated (Stata 17.0). Heterogeneity and bias were evaluated via subgroup/sensitivity analyses and Deeks’ test. Results: 13 studies (4,756 patients) were included. The training set showed pooled SEN 0.90(95%CI:0.81-0.95), SPE 0.82(95%CI:0.72-0.89), and AUC 0.92(95%CI:0.90-0.94). The validation set had SEN 0.74(95%CI:0.62-0.83), SPE 0.80(95%CI:0.70-0.87), and AUC 0.84(95%CI:0.80-0.87). Heterogeneity (I²> 75%) stemmed from sample size and country differences (both P<0.05). Fagan’s plot indicated radiomics increased HER2 overexpression post-test probability to 48%-56% (from 20% pre-test) and reduced negative results to 3%-8%. Sensitivity analysis confirmed robustness; no publication bias was detected (P>0.05). Conclusion: MRI Radiomics correlates with HER2 overexpression in breast cancer and demonstrates high diagnostic accuracy, offering a novel approach for HER2 evaluation.