Differential diagnosis of benign and lung adenocarcinoma presenting as larger solid nodules and masses based on multiscale CT radiomics DOI Creative Commons

Jiayue Xie,

Yifan He,

Siyu Che

et al.

PLoS ONE, Journal Year: 2024, Volume and Issue: 19(10), P. e0309033 - e0309033

Published: Oct. 4, 2024

Purpose To develop a better radiomic model for the differential diagnosis of benign and lung adenocarcinoma lesions presenting as larger solid nodules masses based on multiscale computed tomography (CT) radiomics. Materials methods This retrospective study enrolled 205 patients with from Center 1 between January 2010 February 2022 2 2019 2022. After applying inclusion exclusion criteria, we retrospectively 165 two centers assigned them to training dataset (n = 115) or test 50). Radiomics features were extracted volumes interest CT images. A gradient boosting decision tree (GBDT) was used data dimensionality reduction perform final feature selection. Four models developed using clinical data, conventional imaging radiomics features, namely, image (CIM), plain (PRM), enhanced (ERM) combined (CM). Model performance evaluated determine best identifying masses. Results In dataset, areas under curve (AUCs) CIM, PRM, ERM, CM 0.718, 0.806, 0.819, 0.917, respectively. The diagnostic capability ERM than that PRM CIM. optimal. Intermediate junior radiologists respiratory physicians achieved improved obviously results model. senior showed slight after Conclusion may have potential be noninvasive tool

Language: Английский

Differential diagnosis of benign and lung adenocarcinoma presenting as larger solid nodules and masses based on multiscale CT radiomics DOI Creative Commons

Jiayue Xie,

Yifan He,

Siyu Che

et al.

PLoS ONE, Journal Year: 2024, Volume and Issue: 19(10), P. e0309033 - e0309033

Published: Oct. 4, 2024

Purpose To develop a better radiomic model for the differential diagnosis of benign and lung adenocarcinoma lesions presenting as larger solid nodules masses based on multiscale computed tomography (CT) radiomics. Materials methods This retrospective study enrolled 205 patients with from Center 1 between January 2010 February 2022 2 2019 2022. After applying inclusion exclusion criteria, we retrospectively 165 two centers assigned them to training dataset (n = 115) or test 50). Radiomics features were extracted volumes interest CT images. A gradient boosting decision tree (GBDT) was used data dimensionality reduction perform final feature selection. Four models developed using clinical data, conventional imaging radiomics features, namely, image (CIM), plain (PRM), enhanced (ERM) combined (CM). Model performance evaluated determine best identifying masses. Results In dataset, areas under curve (AUCs) CIM, PRM, ERM, CM 0.718, 0.806, 0.819, 0.917, respectively. The diagnostic capability ERM than that PRM CIM. optimal. Intermediate junior radiologists respiratory physicians achieved improved obviously results model. senior showed slight after Conclusion may have potential be noninvasive tool

Language: Английский

Citations

0