Chengshi Hou, Fang Wang, Martin Prince, Xin Yang, Wenjian Wang, Jing Ye, Lei Chen, Xianfu Luo
{"title":"CT-based liver peritumoural radiomics features predict hepatic metastases sources as gastrointestinal or non-gastrointestinal.","authors":"Chengshi Hou, Fang Wang, Martin Prince, Xin Yang, Wenjian Wang, Jing Ye, Lei Chen, Xianfu Luo","doi":"10.1093/bjr/tqae248","DOIUrl":null,"url":null,"abstract":"<p><strong>Objectives: </strong>To investigate the feasibility of radiomics models for predicting the source of hepatic metastases from gastrointestinal (GI) vs non-gastrointestinal (non-GI) primary tumours on contrast-enhanced CT (CECT).</p><p><strong>Methods: </strong>Three hundred and forty-seven patients with liver metastases (180 from GI and 167 from non-GI) and abdominal CECT including arterial, portal venous, and delayed phases were divided into training (221) and validation (96) sets at a ratio of 7:3 and an independent testing set (30). Radiomics features were extracted from volumes of interest (VOIs) including tumoural (Vtc) and peritumoural (Vpt) regions on CECT. Optimal radiomics features were used in logistic regression models using receiver operating curve (ROC) analysis to evaluate the diagnostic efficiency.</p><p><strong>Results: </strong>The best single-phase model was a venous phase peritumoural VOI with 11 features. Area under the curve (AUC), sensitivity, and specificity were 0.817, 0.740, and 0.761, respectively in the validation set. While the best arterial phase tumoural VOI gave an AUC of 0.677 in the validation set. For the combined models, peritumoural VOI in arterial and venous phases (15 features) achieved the best prediction performance with an AUC of 0.926 in the validation set and 0.884 in the testing set.</p><p><strong>Conclusion: </strong>Liver peritumoural radiomics features extracted from CECT were able to identify the source of hepatic metastases as GI vs non-GI.</p><p><strong>Advances in knowledge: </strong>Peritumoural radiomics features showed a correlation with source of liver metastases. The radiomics features from liver peritumoural arterial and venous phases CT were promising in differentiating the source of hepatic metastases from GI vs non-GI primary tumours.</p>","PeriodicalId":9306,"journal":{"name":"British Journal of Radiology","volume":" ","pages":"458-468"},"PeriodicalIF":1.8000,"publicationDate":"2025-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"British Journal of Radiology","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1093/bjr/tqae248","RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING","Score":null,"Total":0}
引用次数: 0
Abstract
Objectives: To investigate the feasibility of radiomics models for predicting the source of hepatic metastases from gastrointestinal (GI) vs non-gastrointestinal (non-GI) primary tumours on contrast-enhanced CT (CECT).
Methods: Three hundred and forty-seven patients with liver metastases (180 from GI and 167 from non-GI) and abdominal CECT including arterial, portal venous, and delayed phases were divided into training (221) and validation (96) sets at a ratio of 7:3 and an independent testing set (30). Radiomics features were extracted from volumes of interest (VOIs) including tumoural (Vtc) and peritumoural (Vpt) regions on CECT. Optimal radiomics features were used in logistic regression models using receiver operating curve (ROC) analysis to evaluate the diagnostic efficiency.
Results: The best single-phase model was a venous phase peritumoural VOI with 11 features. Area under the curve (AUC), sensitivity, and specificity were 0.817, 0.740, and 0.761, respectively in the validation set. While the best arterial phase tumoural VOI gave an AUC of 0.677 in the validation set. For the combined models, peritumoural VOI in arterial and venous phases (15 features) achieved the best prediction performance with an AUC of 0.926 in the validation set and 0.884 in the testing set.
Conclusion: Liver peritumoural radiomics features extracted from CECT were able to identify the source of hepatic metastases as GI vs non-GI.
Advances in knowledge: Peritumoural radiomics features showed a correlation with source of liver metastases. The radiomics features from liver peritumoural arterial and venous phases CT were promising in differentiating the source of hepatic metastases from GI vs non-GI primary tumours.
期刊介绍:
BJR is the international research journal of the British Institute of Radiology and is the oldest scientific journal in the field of radiology and related sciences.
Dating back to 1896, BJR’s history is radiology’s history, and the journal has featured some landmark papers such as the first description of Computed Tomography "Computerized transverse axial tomography" by Godfrey Hounsfield in 1973. A valuable historical resource, the complete BJR archive has been digitized from 1896.
Quick Facts:
- 2015 Impact Factor – 1.840
- Receipt to first decision – average of 6 weeks
- Acceptance to online publication – average of 3 weeks
- ISSN: 0007-1285
- eISSN: 1748-880X
Open Access option