Table 3.
List of included articles on RFs derived from other imaging modalities than MRI. In the second column # are the number of patients enrolled retrospectively (R) or prospectively (P). In the fourth column the volume of interest (VOI) is presented accompanied by the type of segmentation in brackets M = manual, SA = semiautomatic and A = fully automatic. The last column contains information on validation. “e” stands for external validation and “I” for internal. The number stands for the number of cohorts used. 2 means one cohort for development and one for testing.
| Study | # | Imaging Modality | VOI (Segmentation) | Endpoint(s) | Results | Validation |
|---|---|---|---|---|---|---|
| Prostate | ||||||
| Zhang et al. 47 | 113 (R) | TRUS: B-mode, Sono- elastography |
Prostate (M) | PCa detection | Multimodal feature (4 RFs) learning. | i (2) |
| Wildboer et al. 134. | 50 (R) | TRUS: B-mode, contrast enhanced US |
Prostate (A) | PCa detection, GS | Multiparametric classifier (n=14). | i (CV) |
| Wu et al. 135 | 132 & 5 videos (R) |
TRUS: B-mode |
Prostate (A&M) | Prostate segmentation | Prostate segmentation framework utilizing speckle-induced texture features. | i (2) |
| Huang et al. 136 | 342 (R) | TRUS (M) | Rectangle around the biopsy core | PCa detection | RF for a support vector machine classifier. | i (CV) |
| Osman et al. 133 | 342 (R) | CT | Prostate (M) | GS, risk group discrimination | Radiomics classifier. | i (2, CV) |
| Tanadini-Lang et al. 128 | 41 (R) | CT perfusion | Prostate (M) | GS, risk group discrimination | Single and combined use of RF and conventional CT perfusion parameters. | i (CV) |
| Bosetti et al. 129 | 31 (R) | Cone-beam CT | Prostate (M) | Tumor stage, GS, PSA level, risk group discrimination, BCR | Histogram-based Energy and Kurtosis and a shape-based feature predict BCR and high risk. | i (CV) |
| Toxicity | ||||||
| Mostafei et al. 130 | 64 (P) | CT | Pre-treatment Rectal-& Bladder wall (M) |
RT toxicity GI/GU ≥ grade 1 CTCAEv4.03 |
Cystitis: clinical-radiomics (n=4) model. Proctitis: radiomics (n=3) model. | i (CV) |
| Lymph nodes | ||||||
| Peeken et al. 131 | 80 (R) | Contrast-enhanced CT from PSMA PET/CT scans | Lymph nodes (M) | Lymph node metastasis | Radiomics model significantly outperformed all conventional CT parameters. | i (CV, 2) |
| Bone metastases | ||||||
| Acar et al. 132 | 75 (R) | CT from PSMA PET/CT | Bone metastases | Discrimination of bone metastases that responded after treatment | Weighted k-nearest neighborhood algorithm. | i (CV) |
Abbreviations: CT = computed tomography, CV= cross-validation, GI=gastrointestinal, GS = Gleason score, GU=genitourinary; PCa = prostate cancer; QSZHGE= quantization algorithm + short zones high gray-level emphasis, TRUS = Transrectal Ultrasound