Table 2.

Table summarizing the main studies regarding Radiomics applied to ovarian cancer.

Study	N Patients	Endpoint	Types of Evaluation	Model Performance	Imaging Modality	Features Selected	Nature of Study
Zhang H. et al., Eur. Rad. 2019 [26]	Validation 195 Testing 85	Benign vs. Malignant OEC type I vs. type II	LOO cross-validation Indipendent testing	End. 1 AUC: 0.97 End 1 AUC: 0.85 End 2 AUC: 0.96 End 2 AUC: 0.82	MRI	End. 1: 84 End. 2: 56	Monocentric Restrospective
Song X.L. et al., Eur. Rad. 2021 [27]	Training 72 Validation 32	Benign vs. Borderline Benign vs. Malignant Borderline vs. Malignant	2-class classification	End 1 AUC: 0.89 End 2 AUC: 0.86 End 3 AUC: 0.89	MRI	End. 1: 51 End. 2: 23 End. 3: 18	Monocentric Prospective
Meier A. et al., Abdom. Radiol. 2019 [28]	Total 88	Assosiation Survival and texture heterogeneity	Inter-site texture heterogeneity	p < 0.05	CT	3	Monocentric Restrospective
Lu H. et al., Nat. Commun 2019 [29]	Total 364	Survival	Radiomic prognostic vector	HR > 3.83	CT	4	Multicentric Restrospective
Himoto Y. et al., JCO Precis. Oncol. 2019 [30]	Total 75	Time to off-treatment	Intra-site texture heterogeneity Inter-site texture heterogeneity	p < 0.05 HR: 0.88 HR: 1.19	CT	7	Monocentric Restrospective
Danala. et al., Acad. Radiol 2017 [31]	Total 91	Early prediction treatment response	Delta Radiomics Fusion models	AUC: 0.77 AUC: 0.81–0.82	CT	24	Monocentric Restrospective

LOO, leave-one-out; AUC, area under the curve; HR, hazard ratio; MRI, magnetic resonance imaging; CT, computed tomography.