Page 36 - ARNM-3-2

Advances in Radiotherapy
& Nuclear Medicine Radiomics for gastric cancer

Table 2. Summary of studies on radiomics for distinguishing tumor‑node‑metastasis stage in gastric cancer
Authors Year Image Purpose Sample Features Result
Chen et al. 35 2019 MRI Prediction of LNM in 146 (Multi-center) Radiomics+clinical The specificity, sensitivity, and
patients with advanced accuracy were 0.846, 0.853, and 0.851,
GC respectively, in the internal validation
cohort, and 0.714, 0.952, and 0.893,
respectively, in the external validation
cohort
Jiang et al. 36 2019 CT Radiomic feature-based 1,689 (Multi-center) Radiomics+clinical The radiomic models exhibited good
CT to predict LNM in discriminability for LN staging in
GC the training, internal, and external
validation cohorts
Feng et al. 37 2019 CT Prediction of LNM in 490 (Single center) Radiomics+clinical In the training data, the predicted
patients with GC based AUC of LN+was 0.824 and the
on CT radiomics predicted AUC of the test data was
0.764
Wang et al. 38 2020 CT To investigate the value 244 (Single center) Radiomics Arterial phase-based radiomic model
of CT-based radiomics exhibited an AUC of 0.899 in the
in differentiating stage training cohort and 0.825 in the test
T2 and stage T3/4 GC cohort
Gao et al. 39 2020 CT Prediction of LNM in 463 (Single center) Radiomics+clinical A radiomic model that incorporated
early GC radiomic signature and CT-reported
LN status exhibited a good
discrimination in the training cohort
(AUC=0.91) and testing cohort
(AUC=0.89)
Gao et al. 40 2020 CT Prediction of LNM in 768 (Multi-center) Radiomics+clinical A radiomic model that incorporated
GC radiomic signature, serum CA72-4,
and CT-reported LN status exhibited
a good discrimination in the training
cohort (AUC=0.92) and validation
cohort (AUC=0.86)
Wang et al. 41 2020 CT Prediction of LNM in 274 (Single center) Radiomics+clinical The nomogram consisted of radiomic
GC scores and the CT-reported LN status
exhibited excellent discrimination
in the training and test cohorts with
AUCs of 0.886 and 0.881, respectively
Yardımcı et al. 42 2020 CT To predict the clinical T 114 (Single center) Radiomics Based on the texture features, the
and N stages and tumor differential ability of T stage, N stage,
grade of patients with and tumor grade was 90.4%, 81.6%,
GC before pre-operative and 64.5%, respectively
NAC
Yang et al. 43 2020 CT Prediction of LNM in 170 (Single center) Radiomics The radiomic-clinicopathologic model
GC (training cohort=0.9432±0.0129,
validation cohort=0.8764±0.0322)
showed a good discrimination
capability
Wang et al. 44 2021 CT Identification of no. 10 515 (Single center) Radiomics+clinical CT-based radiomic nomogram
LNs status in advanced yielded classification accuracy,
proximal GC with AUCs of 0.896 and 0.814 in
the training and validation cohort,
respectively.
Sun et al. 45 2021 CT Prediction of LNM 1,618 (Multi-center) Radiomics+clinical The radiomic nomograms revealed
status based on good prediction performances, with
pre-operative CT images AUCs of 0.716 – 0.871 in the training
cohort and 0.678 – 0.768 in the
external validation cohort
(Cont’d...)

Volume 3 Issue 2 (2025) 28 doi: 10.36922/arnm.8350