Differentiating lymphoma from lymphoid inflammatory hyperplasia using 18 F-FDG PET/CT radiomics combined with clinical features
10.19405/j.cnki.issn1000-1492.2025.05.024
- Author:
Liang Xie
1
;
Jialin Qin
1
;
Ruixue Wu
2
;
Chunfeng Xiang
3
;
Pengfei Fang
2
;
Chenfeng Shou
4
;
Hong Chen
5
;
Xiaoxi Pang
1
Author Information
1. Dept of Nuclear Medicine , The Second Afiliated Hospital of Anhui Medical University , Hefei 230601
2. School of Basic Medicine , Anhui Medical University , Hefei 230032
3. Dept of Radiology , Dazhou Central Hospital , Dazhou 635000
4. Dept of Emergency Surgery , The Second Afiliated Hospital of Anhui Medical University , Hefei 230601
5. School of Second Clinical Medical , Anhui Medical University , Hefei 230032
- Publication Type:Journal Article
- Keywords:
radiomics;
lymphoma;
lymphoid inflammatory hyperplasia;
PET/CT;
18 F_FDG;
nuclear medical diagnostics
- From:
Acta Universitatis Medicinalis Anhui
2025;60(5):954-963
- CountryChina
- Language:Chinese
-
Abstract:
Objective :To develop and to validate a combined model integrating18F-FDG PET/CT radiomics with clinical features to distinguish between lymphoma and lymphoid inflammatory hyperplasia.
Methods :A retrospective study was conducted on a cohort of 232 patients diagnosed with lymphoma or lymphoid inflammatory hyperplasia. Comparative analyses of clinical and traditional imaging indicators were performed to identify inter-group differences. The clinical features were delineated and extracted using medical software including 3D-Slicer and Lifex. Selection of the features was performed to construct a PET/CT-based radiomics Logistic model, with a combined model integrating PET/CT with clinical features then used to evaluate the discriminative efficacy of these models.
Results:Analysis of inter-group differences indicated that age, CTmean, and metabolic tumor volume(MTV)were effective for differentiating between lymphoma and lymphoid inflammatory hyperplasia(P<0.05). The PET/CT-based radiomics Logistic model differentiated between lymphoma and lymphoid inflammatory hyperplasia, with an area under curve(AUC) of 0.924(95%CI: 0.884-0.960) and 0.863(95%CI: 0.774-0.939) in the training and testing cohorts, respectively. The integrated Logistic model that combined PET/CT-based radiomics with clinical features to distinguish between lymphoma and lymphoid inflammatory hyperplasia achieved an AUC of 0.933(95%CI: 0.889-0.969) in the training cohort and 0.884(95%CI: 0.792-0.964) in the testing cohort. Decision curve analysis(DCA) demonstrated that the integrated model provided the greatest clinical net benefit.
Conclusion :The hybrid model integrating18F-FDG PET/CT radiomics with clinical features shows robust diagnostic efficacy to distinguish between lymphoma and lymphoid inflammatory hyperplasia.
- Full text:2026020516011085466基于~(18)F-FDG_PET_CT影像组学联合临床特征鉴别淋巴瘤及淋巴炎性增生_谢亮.pdf
