From: Beyond diagnosis: is there a role for radiomics in prostate cancer management?
First author [reference number] | Publication year | Country | Treatment | Aim | Design | Sample size | Imaging modality | Main outcome | Potential impact |
---|---|---|---|---|---|---|---|---|---|
Li [88] | 2021 | USA | RP | BCRFS prediction | Retrospective, multicentre | 198 | MRI (radiomics) | A prognostic nomogram, incorporating pre-operative bpMRI radiomics features and clinicopathologic parameters outperformed CAPRAs score for BCRFS prediction (C-index 0.79 versus 0.68) | Identifying patients at low risk of BCR who might defer additional therapy |
Fernandes [91] | 2018 | The Netherlands | RT | 5-year BCR prediction | Retrospective, single-centre | 120 | MRI (radiomics) | LR model using whole-prostate MRI features (AUC 0.63) outperformed both clinical and combined models (AUC 0.51 and 0.56 respectively) | To develop individualised treatment strategies, stratifying patients at risk of BCR |
Kang [93] | 2020 | USA | RP | BCR prediction | Retrospective, multicentre | 28 | F-18 fluciclovine PET/CT (radiomics) | The model combining Haralick texture features with patients’ clinical parameters improved BCR prediction compared to the models including only clinical data and imaging features (AUC 0.94, 0.71, and 0.92 respectively) | Developing a computational methodology to be used as an adjunct tool to improve and standardise the interpretation of F-18 fluciclovine PET/CT in the identification of BCR |
Papp [92] | 2020 | Austria | RP | BCR prediction | Prospective, single-centre | 52 | [68Ga] Ga-PSMA-11 PET/MRI (radiomics) | Supervised predictive model for BCR, including PET/MRI features and clinical data outperformed the standard routine analysis based on PSA, biopsy GS, and TNM staging (diagnostic accuracy 0.89 versus 0.69) | Identifying patients at risk of BCR who might benefit additional therapy |