Table 1.
Final articles with continuous and discrete parameters. Acc and AUC values as well as number of labels were further investigated for articles with diagnosis-oriented tasks
| Author | Year | Number of patients / cases | Healthy cases | Benign cases | Intermediate cases | Malignant cases | Metastases cases | Study design | Tumour entity group | Imaging modality | Radiomic data | Algorithm | Task | Model | Applied metric | Outcome label | Diagnosis-oriented | Acc | AUC | Number of labels |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Bandyopadhyay et al | 2019 | 150 | 0 | 0 | 0 | 150 | 0 | Retrospective | Bone tumours | X-ray | No | Supervised | Classification | SVM, decision tree | acc, sens, Dice | Histopathological grading, staging | ✓ | 0.85 | 2 | |
| Banerjee et al | 2018 | 21 | 0 | 0 | 0 | 21 | 0 | Retrospective | Soft tissue tumours | MRI | No | Supervised | Classification | AlexNet | acc, AUC, sens, spec | Tumour entities | ✓ | 0.85 | 2 | |
| Chianca et al | 2021 | 146 | 0 | 49 | 0 | 40 | 57 | Retrospective | Bone tumours | MRI | Yes | Supervised | Classification | LogitBoost, SVM | AUC, sens, spec, acc | Malignancy | ✓ | 0.90 | 2 | |
| Do et al | 2021 | 1576 | 381 | 1061 | 0 | 134 | 0 | Retrospective | Bone tumours | X-ray | No | Supervised | Classification, segmentation | UNet | acc, IoU | Segmented tumour, tumour entities | ✓ | 0.99 | 3 | |
| Dufau et al | 2019 | 69 | 0 | 0 | 0 | 69 | 0 | Retrospective | Bone tumours | MRI | Yes | Supervised | Classification | SVM | AUC, sens, spec | Chemotherapy response assessment | × | |||
| Eweje et al | 2021 | 1060 | 0 | 582 | 0 | 478 | 0 | Retrospective | Bone tumours | MRI | No | Supervised | Classification | Efficient-Net, logistic regression | acc, sens, spec, AUC | Malignancy | ✓ | 0.79 | 2 | |
| Fields et al | 2021 | 128 | 0 | 36 | 0 | 92 | 0 | Retrospective | Soft tissue tumours | MRI | Yes | Supervised | Classification | Adaboost, random forest | AUC, sens, spec | Malignancy | ✓ | 0.77 | 2 | |
| Gao et al | 2021 | 30 | 0 | 0 | 0 | 30 | 0 | Prospective | Soft tissue tumours | MRI | No | Supervised | Classification | VGG19 | sens, spec, acc | Radiotherapy response assessment | × | |||
| Gao et al | 2020 | 30 | 0 | 0 | 0 | 30 | 0 | Prospective | Soft tissue tumours | MRI | Yes | Supervised | Classification | SVM, logistic regression | AUC | Radiotherapy response assessment | × | |||
| García-Gómez et al | 2004 | 430 | 0 | 267 | 0 | 163 | 0 | Retrospective | Soft tissue tumours | MRI | No | Supervised | Classification | K-nearest neighbour, SVM | sens, spec | Malignancy | ✓ | 0.90 | 2 | |
| Gitto et al | 2020 | 58 | 0 | 0 | 0 | 58 | 0 | Retrospective | Bone tumours | MRI | Yes | Supervised | Classification | LogitBoost | acc, AUC | Histopathological grading | ✓ | 0.75 | 0.78 | 2 |
| Glass et al | 1998 | 43 | 0 | 0 | 0 | 43 | 0 | Retrospective | Bone tumours | MRI | No | Unsupervised | Segmentation | Neural network | acc, sens, spec | Chemotherapy response assessment | × | |||
| He et al | 2020 | 1356 | 0 | 679 | 0 | 360 | 317 | Retrospective | Bone tumours | X-ray | No | Supervised | Classification | Efficient-Net | AUC, sens, spec, acc | Malignancy | ✓ | 0.73 | 2 | |
| Holbrook et al | 2020 | 79 | 0 | 0 | 0 | 79 | 0 | Unknown | Soft tissue tumours | MRI | Yes | Supervised | Segmentation | SVM, neural network | Dice, AUC | Segmented tumour | × | |||
| Hu et al | 2021 | 160 | 0 | 90 | 0 | 70 | 0 | Retrospective | Soft tissue tumours | MRI | Yes | Supervised | Classification | Least absolute shrinkage and selection operator | AUC, sens, spec, acc | Malignancy | ✓ | 0.92 | 0.96 | 2 |
| Hu et al | 2014 | 141 | 0 | 71 | 0 | 70 | 0 | Unknown | Bone tumours | X-ray | No | Supervised | Classification | SVM | acc, AUC, sens, spec | Tumour occurrence | ✓ | 0.96 | 2 | |
| Huang et al | 2020 | 12 | 0 | 0 | 0 | 12 | 0 | Prospective | Bone tumours | MRI | No | Supervised | Classification | Random forest | AUC, sens, spec, acc | Chemotherapy response assessment | × | |||
| Huang et al | 2017 | 23 | 0 | 0 | 0 | 23 | 0 | Unknown | Bone tumours | CT | No | Supervised | Segmentation | VGG16 | Dice score | Segmented tumour | × | |||
| Juntu et al | 2010 | 135 | 0 | 86 | 0 | 49 | 0 | Unknown | Soft tissue tumours | MRI | No | Supervised | Classification | SVM, neural network, decision tree | AUC, sens, spec, acc | Malignancy | ✓ | 0.93 | 2 | |
| Leporq et al | 2020 | 81 | 0 | 40 | 0 | 41 | 0 | Retrospective | Soft tissue tumours | MRI | Yes | Supervised | Classification | SVM | AUC, sens, spec, acc | Malignancy | ✓ | 0.95 | 0.96 | 2 |
| Li et al | 2019 | 210 | 0 | 154 | 0 | 56 | 0 | Retrospective | Bone tumours | MRI | Yes | Supervised | Classification | SVM | AUC, sens, spec, acc | Tumour entities | ✓ | 0.87 | 2 | |
| Liu et al | 2021 | 643 | 0 | 392 | 93 | 158 | 0 | Retrospective | Bone tumours | X-ray | No | Supervised | Classification | XGBoost, Inception V3 | AUC, sens, spec, acc | Malignancy | ✓ | 0.87 | 3 | |
| Pan et al | 2021 | 796 | 0 | 412 | 169 | 215 | 0 | Retrospective | Bone tumours | X-ray | No | Supervised | Classification | Random forest | AUC, acc | Malignancy | ✓ | 0.95 | 0.97 | 3 |
| Peeken et al | 2019 | 221 | 0 | 221 | 0 | 0 | 0 | Retrospective | Soft tissue tumours | CT | Yes | Supervised | Classification | Random forest | AUC, Dice | Histopathological grading | ✓ | 0.64 | 2 | |
| Peeken et al | 2018 | 136 | 0 | 0 | 0 | 136 | 0 | Retrospective | Soft tissue tumours | MRI, CT | No | Supervised | Classification | Random forest | AUC, sens, spec, acc | Prognosis | × | |||
| Reinus et al | 1994 | 709 | 0 | 492 | 0 | 217 | 0 | Retrospective | Bone tumours | X-ray | No | Supervised | Classification | Neural network | acc | Malignancy | ✓ | 0.85 | 2 | |
| Shen et al | 2018 | 36 | 0 | 15 | 0 | 21 | 0 | Unknown | Bone tumours | X-ray | No | Supervised | Classification | Random forest, SVM | AUC, sens, spec, acc | Malignancy | ✓ | 0.85 | 0.94 | 2 |
| Terunuma et al | 2018 | 1 | N/A | N/A | N/A | N/A | N/A | Retrospective | Bone tumours | X-ray | No | Supervised | Object detection, segmentation | SegNet | Jaccard index | Segmented tumour | × | |||
| von Schacky et al | 2021 | 934 | 0 | 623 | 0 | 311 | 0 | Retrospective | Bone tumours | X-ray | No | Supervised | Object detection, segmentation, classification | Mask-RCNN | acc, sens, spec, IoU, Dice | Malignancy | × | |||
| Vos et al | 2019 | 116 | 0 | 58 | 0 | 58 | 0 | Retrospective | Soft tissue tumours | MRI | Yes | Supervised | Classification | SVM, random forest | AUC, sens, spec | Tumour entities | ✓ | 0.89 | 2 | |
| Wang et al | 2021 | 227 | 0 | 147 | 0 | 80 | 0 | Retrospective | Bone tumours | US | No | Supervised | Classification | VGG16 | acc, sens, spec, AUC | Malignancy | ✓ | 0.79 | 0.91 | 2 |
| Wang et al | 2020 | 206 | 0 | 105 | 0 | 93 | 8 | Retrospective | Soft tissue tumours | MRI | Yes | Supervised | Classification | SVM, generalised linear models, random forest | AUC, sens, spec, acc | Malignancy | ✓ | 0.86 | 0.92 | 2 |
| Yin et al | 2019 | 120 | 0 | 0 | 30 | 54 | 36 | Retrospective | Bone tumours | MRI | Yes | Supervised | Classification | Random forest | AUC, acc | Segmented tumour, tumour entities | ✓ | 0.71 | 0.77 | 3 |
| Yin et al | 2019 | 95 | 0 | 0 | 42 | 53 | 0 | Retrospective | Bone tumours | CT | Yes | Supervised | Classification | Random forest | acc, AUC | Tumour entities | ✓ | 0.90 | 0.98 | 2 |
| Yin et al | 2021 | 795 | 0 | 215 | 0 | 399 | 181 | Retrospective | Bone tumours | CT | Yes | Supervised | Classification | Random forest | AUC, acc | Tumour entities | ✓ | 0.88 | 0.93 | 2 |
| Zhang et al | 2020 | 51 | N/A | N/A | N/A | N/A | N/A | Retrospective | Soft tissue tumours | MRI, CT | No | Supervised | Classification | Inception-v3 | acc, AUC | Histopathological grading | ✓ | 0.86 | 0.97 | 3 |
| Zhang et al | 2019 | 35 | 0 | 0 | 0 | 35 | 0 | Retrospective | Soft tissue tumours | MRI | Yes | Supervised | Classification | Random forest, SVM | AUC, sens, spec, acc | Histopathological grading | ✓ | 0.88 | 0.92 | 2 |
| Zhang et al | 2018 | 23 | 0 | 0 | 0 | 23 | 0 | Unknown | Bone tumours | CT | No | Supervised | Segmentation | ResNet-50 | Dice, sens | Segmented tumour | × |
SVM support vector machine, IoU intersection over union N/A not assessed