Purpose: In the last years, it has emerged the unmet need for imaging technologies able to accurately describe the biological and clinical heterogeneity of glioma, assess responsiveness to therapy, and predict patient outcomes. The increasing number of additional imaging parameters primarily derived from advanced Magnetic Resonance Imaging (MRI) and Positron Emission Tomography (PET), the technical developments of hybrid PET/MRI scanners, and the use of artificial intelligence might be able to fill this gap. In the present narrative review, we analyzed the potential of radiomics features extracted from stand-alone PET/MRI or combined PET and MRI scanners in the assessment of glioma. Methods: PubMed, Embase, and Web of Science were searched for articles published up to January 2021 using the following keywords: “PET/MRI AND glioma”, “PET/MRI AND glioblastoma”, “PET/MRI AND radiomics AND glioma”, “PET/MRI AND radiomics AND glioblastoma”. We included only original articles published in English reporting the use of PET/MRI or PET/MRI-derived radiomics features (extracted either from PET or MRI scanner or hybrid PET/MRI device) in patients with glioma. Results: Published data suggest that hybrid PET/MRI may be superior to MRI-alone in differentiating low-grade vs. high-grade glioma, in detecting malignant transformation of LGG, and in the differential diagnosis between true recurrences from post-treatment changes. Moreover, PET/MRI could add helpful information for radiotherapy and surgery planning. Also, the rapid advancement in computer-based image analysis including artificial intelligence and radiomics has extended to PET imaging in the last years, with promising results particularly in the setting of preoperative imaging and in the post-treatment setting. Conclusions: Published literature suggests a high potential for PET/MRI-derived radiomics. However, these results are based on small datasets and need to be further investigated in large prospective studies. Multidisciplinary efforts and multicenter collaborations will be essential to achieve these aims in the next future.
Emerging applications of imaging in glioma: focus on PET/MRI and radiomics
Bauckneht M.;Donegani M. I.;
2021-01-01
Abstract
Purpose: In the last years, it has emerged the unmet need for imaging technologies able to accurately describe the biological and clinical heterogeneity of glioma, assess responsiveness to therapy, and predict patient outcomes. The increasing number of additional imaging parameters primarily derived from advanced Magnetic Resonance Imaging (MRI) and Positron Emission Tomography (PET), the technical developments of hybrid PET/MRI scanners, and the use of artificial intelligence might be able to fill this gap. In the present narrative review, we analyzed the potential of radiomics features extracted from stand-alone PET/MRI or combined PET and MRI scanners in the assessment of glioma. Methods: PubMed, Embase, and Web of Science were searched for articles published up to January 2021 using the following keywords: “PET/MRI AND glioma”, “PET/MRI AND glioblastoma”, “PET/MRI AND radiomics AND glioma”, “PET/MRI AND radiomics AND glioblastoma”. We included only original articles published in English reporting the use of PET/MRI or PET/MRI-derived radiomics features (extracted either from PET or MRI scanner or hybrid PET/MRI device) in patients with glioma. Results: Published data suggest that hybrid PET/MRI may be superior to MRI-alone in differentiating low-grade vs. high-grade glioma, in detecting malignant transformation of LGG, and in the differential diagnosis between true recurrences from post-treatment changes. Moreover, PET/MRI could add helpful information for radiotherapy and surgery planning. Also, the rapid advancement in computer-based image analysis including artificial intelligence and radiomics has extended to PET imaging in the last years, with promising results particularly in the setting of preoperative imaging and in the post-treatment setting. Conclusions: Published literature suggests a high potential for PET/MRI-derived radiomics. However, these results are based on small datasets and need to be further investigated in large prospective studies. Multidisciplinary efforts and multicenter collaborations will be essential to achieve these aims in the next future.
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
