Table 1.

Summary of some radiomic studies using both structural and functional images

Study	Study design	Modality	Population	Features type	Features selection	Model algorithm	Type of validation	Outcome
Du et al. (2020)	Retrospective Single-center	PET/CT	77 Tuberculosis 79 Lung cancers	Clinical Semantic Handcrafted radiomic	LASSO	LR	Leave one out Without external validation	Diagnosis
Sibille et al. (2020)	Retrospective Single-center	PET/CT	302 lung cancer 327 lymphoma	Deep radiomic	CNN	CNN	Leave one out Without external validation	Diagnosis
Kang et al. (2019)	Retrospective Single-center	PET/CT	157 malignant 111 benign patients	Clinical Handcrafted radiomic	LASSO	LR	Bootstrapping validation Without external validation	Diagnosis
Han et al. (2020)	Retrospective Single-center	PET/CT	867 adenocarcinomas 552 SCCs	Handcrafted radiomic Deep radiomic	Ten feature selection methods	10 ML models and the VGG16	Leave one out Without external validation	Diagnosis
Kirienko et al. (2018)	Retrospective Single-center	PET/CT	534 Lung lesions	Handcrafted radiomic	–	LDA	Leave one out Without external validation	Primary or metastatic lung lesions
Shao et al. (2020)	Retrospective Single-center	PET/CT	91 GGNs	Semantic Handcrafted radiomic	LASSO	LR	Bootstrapping validation Without external validation	Lepidic or acinar-papillary growth
Zhang et al. (2020)	Retrospective Single-center	PET/CT	248 NSCLCs	Clinical Handcrafted radiomic	LASSO	LR	Leave one out Without external validation	EGFR mutation
Liu et al. (2020)	Retrospective Single-center	PET/CT	148 Adenocarcinomas	Handcrafted radiomic	RF/LR	Xgboost	Leave one out Without external validation	EGFR mutation
Mu et al. (2020)	Retrospective Multi-center	PET/CT	681 NSCLCs	Deep radiomic	CNN	CNN	Leave one out With external validation	EGFR mutation Treatment response
Yang et al. (2020)	Retrospective Single-center	PET/CT	315 NSCLCs	Clinical Handcrafted radiomic	LASSO	LR	Leave one out Without external validation	Survival
Mu et al. (2019)	Retrospective/ prospective Single-center	PET/CT	194 Stage IIIB-IV NSCLCs	Clinical Handcrafted radiomic	LASSO	LR	Leave one out With external validation	Survival after immunotherapy
Dissaux et al. (2020)	Retrospective Multi-center	PET/CT	87 Early-stage NSCLCs	Clinical Semantic Handcrafted radiomic	Univariate/ Multivariate analysis	Cox	Leave one out With external validation	Local Recurrence after radiotherapy

SCC, squamous cell carcinoma; NSCLC, non-small cell lung cancer; GGN, ground-glass nodule; LASSO, least absolute shrinkage and selection operator; CNN, convolutional neural network; RF, random forest; LR, logistic regression; ML, machine learning; EGFR, epidermal growth factor receptor