Robustness of PET Radiomics Features: Impact of Co-Registration with MRI
Radiomics holds great promise in the field of cancer management. However, the clinical application of radiomics has been hampered by uncertainty about the robustness of the features extracted from the images. Previous studies have reported that radiomics features are sensitive to changes in voxel si...
| Main Authors: | , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-10-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/11/21/10170 |
| _version_ | 1797512781234176000 |
|---|---|
| author | Alessandro Stefano Antonio Leal Selene Richiusa Phan Trang Albert Comelli Viviana Benfante Sebastiano Cosentino Maria G. Sabini Antonino Tuttolomondo Roberto Altieri Francesco Certo Giuseppe Maria Vincenzo Barbagallo Massimo Ippolito Giorgio Russo |
| author_facet | Alessandro Stefano Antonio Leal Selene Richiusa Phan Trang Albert Comelli Viviana Benfante Sebastiano Cosentino Maria G. Sabini Antonino Tuttolomondo Roberto Altieri Francesco Certo Giuseppe Maria Vincenzo Barbagallo Massimo Ippolito Giorgio Russo |
| author_sort | Alessandro Stefano |
| collection | DOAJ |
| description | Radiomics holds great promise in the field of cancer management. However, the clinical application of radiomics has been hampered by uncertainty about the robustness of the features extracted from the images. Previous studies have reported that radiomics features are sensitive to changes in voxel size resampling and interpolation, image perturbation, or slice thickness. This study aims to observe the variability of positron emission tomography (PET) radiomics features under the impact of co-registration with magnetic resonance imaging (MRI) using the difference percentage coefficient, and the Spearman’s correlation coefficient for three groups of images: (i) original PET, (ii) PET after co-registration with T1-weighted MRI and (iii) PET after co-registration with FLAIR MRI. Specifically, seventeen patients with brain cancers undergoing [11C]-Methionine PET were considered. Successively, PET images were co-registered with MRI sequences and 107 features were extracted for each mentioned group of images. The variability analysis revealed that shape features, first-order features and two subgroups of higher-order features possessed a good robustness, unlike the remaining groups of features, which showed large differences in the difference percentage coefficient. Furthermore, using the Spearman’s correlation coefficient, approximately 40% of the selected features differed from the three mentioned groups of images. This is an important consideration for users conducting radiomics studies with image co-registration constraints to avoid errors in cancer diagnosis, prognosis, and clinical outcome prediction. |
| first_indexed | 2024-03-10T06:06:32Z |
| format | Article |
| id | doaj.art-6d61fbbbf8fe4a89b7c6e22effebaaec |
| institution | Directory Open Access Journal |
| issn | 2076-3417 |
| language | English |
| last_indexed | 2024-03-10T06:06:32Z |
| publishDate | 2021-10-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj.art-6d61fbbbf8fe4a89b7c6e22effebaaec2023-11-22T20:28:58ZengMDPI AGApplied Sciences2076-34172021-10-0111211017010.3390/app112110170Robustness of PET Radiomics Features: Impact of Co-Registration with MRIAlessandro Stefano0Antonio Leal1Selene Richiusa2Phan Trang3Albert Comelli4Viviana Benfante5Sebastiano Cosentino6Maria G. Sabini7Antonino Tuttolomondo8Roberto Altieri9Francesco Certo10Giuseppe Maria Vincenzo Barbagallo11Massimo Ippolito12Giorgio Russo13Institute of Molecular Bioimaging and Physiology, National Research Council (IBFM-CNR), 90015 Cefalù, ItalyDepartamento de Fisiología Médica y Biofísica, University de Seville/Instituto de Biomedicina de Sevilla (IBiS), 41013 Seville, SpainInstitute of Molecular Bioimaging and Physiology, National Research Council (IBFM-CNR), 90015 Cefalù, ItalyDepartment of Physics and Astronomy “E. Majorana”, University of Catania, 95124 Catania, ItalyRi.Med Foundation, Via Bandiera 11, 90133 Palermo, ItalyInstitute of Molecular Bioimaging and Physiology, National Research Council (IBFM-CNR), 90015 Cefalù, ItalyNuclear Medicine Department, Cannizzaro Hospital, 95126 Catania, ItalyNuclear Medicine Department, Cannizzaro Hospital, 95126 Catania, ItalyDepartment of Health Promotion, Mother and Child Care, Internal Medicine and Medical Specialties, Molecular and Clinical Medicine, University of Palermo, 90127 Palermo, ItalyNeurosurgical Unit, AOU Policlinico “G. Rodolico-San Marco”, University of Catania, 95123 Catania, ItalyNeurosurgical Unit, AOU Policlinico “G. Rodolico-San Marco”, University of Catania, 95123 Catania, ItalyNeurosurgical Unit, AOU Policlinico “G. Rodolico-San Marco”, University of Catania, 95123 Catania, ItalyNuclear Medicine Department, Cannizzaro Hospital, 95126 Catania, ItalyInstitute of Molecular Bioimaging and Physiology, National Research Council (IBFM-CNR), 90015 Cefalù, ItalyRadiomics holds great promise in the field of cancer management. However, the clinical application of radiomics has been hampered by uncertainty about the robustness of the features extracted from the images. Previous studies have reported that radiomics features are sensitive to changes in voxel size resampling and interpolation, image perturbation, or slice thickness. This study aims to observe the variability of positron emission tomography (PET) radiomics features under the impact of co-registration with magnetic resonance imaging (MRI) using the difference percentage coefficient, and the Spearman’s correlation coefficient for three groups of images: (i) original PET, (ii) PET after co-registration with T1-weighted MRI and (iii) PET after co-registration with FLAIR MRI. Specifically, seventeen patients with brain cancers undergoing [11C]-Methionine PET were considered. Successively, PET images were co-registered with MRI sequences and 107 features were extracted for each mentioned group of images. The variability analysis revealed that shape features, first-order features and two subgroups of higher-order features possessed a good robustness, unlike the remaining groups of features, which showed large differences in the difference percentage coefficient. Furthermore, using the Spearman’s correlation coefficient, approximately 40% of the selected features differed from the three mentioned groups of images. This is an important consideration for users conducting radiomics studies with image co-registration constraints to avoid errors in cancer diagnosis, prognosis, and clinical outcome prediction.https://www.mdpi.com/2076-3417/11/21/10170radiomics feature robustnessimaging quantification[11C]-methionine positron emission tomographyPET/MRI co-registration |
| spellingShingle | Alessandro Stefano Antonio Leal Selene Richiusa Phan Trang Albert Comelli Viviana Benfante Sebastiano Cosentino Maria G. Sabini Antonino Tuttolomondo Roberto Altieri Francesco Certo Giuseppe Maria Vincenzo Barbagallo Massimo Ippolito Giorgio Russo Robustness of PET Radiomics Features: Impact of Co-Registration with MRI Applied Sciences radiomics feature robustness imaging quantification [11C]-methionine positron emission tomography PET/MRI co-registration |
| title | Robustness of PET Radiomics Features: Impact of Co-Registration with MRI |
| title_full | Robustness of PET Radiomics Features: Impact of Co-Registration with MRI |
| title_fullStr | Robustness of PET Radiomics Features: Impact of Co-Registration with MRI |
| title_full_unstemmed | Robustness of PET Radiomics Features: Impact of Co-Registration with MRI |
| title_short | Robustness of PET Radiomics Features: Impact of Co-Registration with MRI |
| title_sort | robustness of pet radiomics features impact of co registration with mri |
| topic | radiomics feature robustness imaging quantification [11C]-methionine positron emission tomography PET/MRI co-registration |
| url | https://www.mdpi.com/2076-3417/11/21/10170 |
| work_keys_str_mv | AT alessandrostefano robustnessofpetradiomicsfeaturesimpactofcoregistrationwithmri AT antonioleal robustnessofpetradiomicsfeaturesimpactofcoregistrationwithmri AT selenerichiusa robustnessofpetradiomicsfeaturesimpactofcoregistrationwithmri AT phantrang robustnessofpetradiomicsfeaturesimpactofcoregistrationwithmri AT albertcomelli robustnessofpetradiomicsfeaturesimpactofcoregistrationwithmri AT vivianabenfante robustnessofpetradiomicsfeaturesimpactofcoregistrationwithmri AT sebastianocosentino robustnessofpetradiomicsfeaturesimpactofcoregistrationwithmri AT mariagsabini robustnessofpetradiomicsfeaturesimpactofcoregistrationwithmri AT antoninotuttolomondo robustnessofpetradiomicsfeaturesimpactofcoregistrationwithmri AT robertoaltieri robustnessofpetradiomicsfeaturesimpactofcoregistrationwithmri AT francescocerto robustnessofpetradiomicsfeaturesimpactofcoregistrationwithmri AT giuseppemariavincenzobarbagallo robustnessofpetradiomicsfeaturesimpactofcoregistrationwithmri AT massimoippolito robustnessofpetradiomicsfeaturesimpactofcoregistrationwithmri AT giorgiorusso robustnessofpetradiomicsfeaturesimpactofcoregistrationwithmri |