Preoperative prediction power of radiomics and non-radiomics methods based on MRI for early recurrence in hepatocellular carcinoma: a systemic review and meta-analysis
To compare radiomics and non-radiomics in predicting early recurrence (ER) in patients with hepatocellular carcinoma (HCC) after curative surgery.
Methods
We systematically searched PubMed and Embase databases. Studies with clear reference criteria were selected. Data were extracted and assessed for quality using the quality in prognosis studies tool (QUIPS) by two independent authors. All included radiomics studies underwent radiomics quality score (RQS) assessment. We calculated sensitivity, specificity, positive likelihood ratio (PLR), and negative likelihood ratio (NLR) using random or fixed models with a 95%CI. Forest maps visualized the data, and summary receiver operating characteristic (sROC) curves with the area under the curve (AUC) were generated. Meta-regression and subgroup analyses explored sources of heterogeneity. We compared sensitivity, specificity, PLR, and NLR using the z-test and compared AUC values using the Delong test.
Results
Our meta-analysis included 10 studies comprising 1857 patients. For radiomics, the pooled sensitivity, specificity, AUC of sROC, PLR and NLR were 0.84(95%CI: 0.78–0.89), 0.80(95%CI: 0.75–0.85), 0.89(95%CI: 0.86–0.91), 4.28(95%CI: 3.48–5.27) and 0.20(95%CI: 0.14–0.27), respectively, but with significant heterogeneity (I2 = 60.78% for sensitivity, I2 = 55.79% for specificity) and potential publication bias (P = 0.04). The pooled sensitivity, specificity, AUC of sROC, PLR, NLR for non-radiomics were 0.75(95%CI:0.68–0.81), 0.78(95%CI:0.72–0.83), 0.83(95%CI: 0.80–0.86), 3.45(95%CI: 2.68–4.44) and 0.32(95%CI: 0.24–0.41), respectively. There was no significant heterogeneity in this group (I2 = 0% for sensitivity, I2 = 17.27% for specificity). Radiomics showed higher diagnostic accuracy (AUC: 0.89 vs. 0.83, P = 0.0456), higher sensitivity (0.84 vs. 0.75, P = 0.0385) and lower NLR (0.20 vs. 0.32, P = 0.0287).
Conclusion
The radiomics from preoperative MRI effectively predicts ER of HCC and has higher diagnostic accuracy than non-radiomics. Due to potential publication bias and suboptimal RQS scores in radiomics, these results should be interpreted cautiously.
Preoperative prediction power of radiomics and non-radiomics methods based on MRI for early recurrence in hepatocellular carcinoma: a systemic review and meta-analysis
Authors
Mingjie Lu Chen Wang Yi Zhuo Junjiu Gou Yingfeng Li Jingqi Li Xue Dong
WHO estimates that half of all patients worldwide are non-adherent to their prescribed medication. The consequences of poor adherence can be catastrophic, on both the individual and population level.
Join our expert panel to discover why you need to understand the drivers of non-adherence in your patients, and how you can optimize medication adherence in your clinics to drastically improve patient outcomes.
Watch Dr. Anne Marie Valente present the last year's highlights in pediatric and congenital heart disease in the official ACC.24 Year in Review session.