Table 3.
Selected characteristics of included studies.
| References | Country | Year | Study design |
Patients (% female patients) |
Sample size for diagnostic accuracy | Mean or median age (SD; range), years | Imaging modality | Type of malignancy | AI model (Per-patient/per-node diagnostic output) | Reference standard | Classification criteria |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Coroller et al. (13) | USA | 2016 | Retrospective single-center | 85 (65%) | – | 60.3 | CT | NSCLC | Radiomics (per-patient) | Radiology | B D |
| Parmar et al. (14) | USA | 2018 | Retrospective single-center | 1,194 | – | 68.3 (32–93) | CT | NSCLC | Deep learning (per-patient) | Pathology | A B C |
| Sun et al. (15) | China | 2019 | Retrospective single-center | 385 (68%) | 201 | 53.1 (±12.2) | CT | Lung Cancer | Radiomics (per-patient) | Radiology | A C |
| Ling et al. (16) | China | 2019 | Retrospective multi-center | 229 (31.5%) | 74 | 64 (59–81) | CT | Lung Cancer | Radiomics (per-patient) | Radiology | A |
| Coudray et al. (17) | USA | 2018 | Retrospective single-center | 1,176 | 459 | 61 (51.3–72.8) | CT | NSCLC | Deep learning (per-patient) | Radiology | B C |
| Xu et al. (18) | China | 2019 | Retrospective single-center | 179 (52.8%) | – | 63 (32–93) | CT | NSCLC | Deep learning (per-patient) | Pathology | B D |
| Baldwin et al. (19) | UK | 2020 | Retrospective single-center | 1,337 | 328 | – | CT | Lung Cancer | Deep learning (per-patient) | – | A |
| Schroers et al. (20) | Germany | 2019 | Retrospective single-center | 82 (38%) | 50 | 61.5 (±5.0) | MRI | Lung Cancer | Radiomics (per-patient) | Pathology | A C |
| Wang et al. (21) | China | 2019 | Retrospective single-center | 249 (39.8%) | – | 61.4 (±8.96) | CT | Lung Cancer | Deep learning (per-patient) | Radiology | D |
| Leleu et al. (22) | France | 2020 | Retrospective single-center | 215 (39%) | 72 | 58.6 (±10.3) | CT | Lung Cancer | Radiomics (per-patient) | Pathology | A |
| Ann et al. (23) | USA | 2019 | Prospective multi-center | 262 | 48 | – | CT | NSCLC | Radiomics (per-patient) | Pathology | A B C |
| Cong et al. (24) | China | 2020 | Retrospective single-center | 411 (50.4%) | 141 | 59.62 (24–84) | CT | NSCLC | Radiomics (per-patient) | Radiology | B C D |
| Botta et al. (25) | Italy | 2020 | Retrospective single-center | 270 (38%) | – | 67.4 (61.0–72.6) | CT | NSCLC | Radiomics (per-patient) | Radiology | A B D |
| Wei et al. (26) | USA | 2020 | Retrospective multi-center | 146 (39.7%) | – | 65.72 (± 12.88) | PET/CT | NSCLC | Radiomics (per-node) | Radiology | A B C |
| Khorrami et al. (27) | USA | 2019 | Retrospective single-center | 112 | – | – | CT | NSCLC | Radiomics (per-patient) | Pathology | B D |
| Kirienko et al. (28) | Italy | 2021 | Retrospective single-center | 149 (37.6%) | 73 | 70 (41–84) | PET/CT | Lung Cancer | Radiomics (per-node) | Radiology | B C |
| Rossi et al. (29) | Italy | 2020 | Retrospective single-center | 109 | – | – | CT | NSCLC | Radiomics (per-patient) | Radiology | A B |
| Chai et al. (30) | China | 2021 | Retrospective single-center | 198 (54%) | 402 | 58.1 (± 8.5) | CT | NSCLC | Radiomics (per-node) | Pathology | A B D |
| Wang et al. (31) | China | 2019 | Retrospective single-center | 717 | 386 | — | CT | NSCLC | Radiomics (per-node) | Radiology | B D |
A, Determine whether the patient has lung cancer; B, Determine whether the patient has non-small cell lung cancer; C, Determine whether the patient has malignant lung nodule; D, Determine whether the patient has lymph node metastasis.