본문으로 건너뛰기
← 뒤로

Radiomics-based Machine Learning for Predicting Response to Immune Checkpoint Inhibitors in Hepatocellular Carcinoma: A Systematic Review and Meta-analysis.

Academic radiology 2025 Vol.32(11) p. 6478-6490

Liang WQ, Yang C, Qin LH, Lan MN, Liao PJ

📝 환자 설명용 한 줄

[RATIONALE AND OBJECTIVES] Immune checkpoint inhibitors (ICI) have become a key option for systemic therapy in hepatocellular carcinoma (HCC).

🔬 핵심 임상 통계 (초록에서 자동 추출 — 원문 검증 권장)
  • 95% CI 0.75-0.88

이 논문을 인용하기

BibTeX ↓ RIS ↓
APA Liang WQ, Yang C, et al. (2025). Radiomics-based Machine Learning for Predicting Response to Immune Checkpoint Inhibitors in Hepatocellular Carcinoma: A Systematic Review and Meta-analysis.. Academic radiology, 32(11), 6478-6490. https://doi.org/10.1016/j.acra.2025.07.008
MLA Liang WQ, et al.. "Radiomics-based Machine Learning for Predicting Response to Immune Checkpoint Inhibitors in Hepatocellular Carcinoma: A Systematic Review and Meta-analysis.." Academic radiology, vol. 32, no. 11, 2025, pp. 6478-6490.
PMID 40707262

Abstract

[RATIONALE AND OBJECTIVES] Immune checkpoint inhibitors (ICI) have become a key option for systemic therapy in hepatocellular carcinoma (HCC). Radiomics provides quantitative tumor characteristics that may predict ICI response. This paper assesses the predictive accuracy of radiomics-based machine learning models for ICI treatment response in HCC.

[MATERIALS AND METHODS] We systematically searched PubMed, Embase, Web of Science, and Cochrane Library up to February 14, 2025. The included studies were evaluated for methodological quality and bias risk using three tools: the Prediction model Risk of Bias Assessment Tool (PROBAST), Radiomics Quality Score (RQS), and Methodological Radiomics Score (METRICS). Pooled sensitivity, specificity, positive likelihood ratio (PLR), and negative likelihood ratio (NLR) were calculated using a bivariate random-effects model. Summary receiver operating characteristic (sROC) curves were constructed, with the area under the curve (AUC) calculated to evaluate predictive accuracy. Potential sources of heterogeneity were examined through meta-regression and subgroup analyses, including imaging modalities, modeling variables, region of interest, data sources, and validation method.

[RESULTS] Our analysis incorporated 11 cohorts from 9 studies. The pooled sensitivity, specificity, PLR, NLR, and AUC were 0.83 (95% CI: 0.75-0.88), 0.79 (95% CI: 0.68-0.88), 4.03 (95% CI: 2.57-6.31), 0.22 (95% CI: 0.16-0.30), 0.88 (95% CI: 0.85-0.90), respectively. Meta-regression and subgroup analyses identified several sources of heterogeneity, including imaging modalities, modeling variables, data sources, and validation method.

[CONCLUSION] Radiomics-based machine learning models demonstrate promising efficacy in predicting treatment response to ICI in HCC patients. Nevertheless, further research is warranted to enhance radiomics methodologies and verify these findings in larger, multi-center cohorts.

MeSH Terms

Humans; Liver Neoplasms; Immune Checkpoint Inhibitors; Carcinoma, Hepatocellular; Machine Learning; Treatment Outcome; Radiomics

같은 제1저자의 인용 많은 논문 (1)