Table 1.
Radiomics harmonization in recent multicenter MRI studies.
| Reference | MRI Sequences/ Cancer Type |
Variation Across | Harmonization Techniques | Clinical Question |
|---|---|---|---|---|
| A. Carre et al. (2020) [16] | T1w, T2w-fl1air (brain-glioma) |
3 centers | Intensity-based | Tumor grade classification |
| M. Bologna et al. (2019) [20] | T1w, T2w (brain) |
27 centers | Intensity-based | Feature stability |
| H. Moradmand et al. (2019) [21] | FLAIR, T1, T1C, and T2 (brain-glioblastoma) |
8 centers | Intensity-based | Edema, necrosis, enhancement, and tumor |
| Y. Li et al. (2021) [22] | T1 (brain) |
2 scanners | Both | Scanner effect removal |
| L. J Isaksson et al. (2020) [23] | T2w (prostate cancer) |
1 scanner | Intensity-based | Cancer identification |
| A. Crombe et al. (2020) [24] | T2w (sarcoma) |
3 centers | Intensity-based | Prediction of metastatic-relapse-free survival |
| Chatterjee et al. (2019) [25] | T2-weighted fast spin echo (denoted as T2); T1-weighted fast gradient echo with DCE-MRI and a postcontrast image (post-Gado); and diffusion-weighted MRI (primary uterine adenocarcinoma) |
2 centers | Intensity-based | Lymphovascular space invasion and cancer staging |
| K.A. Wahid et al. (2021) [26] | T2w (head and neck) |
15 HET cohorts-15 HOM- cohorts | Intensity-based | Radiotherapy treatment |
| J. C. Reinhold et al. (2018) [27] | T1-w, T2-w, and FLAIR (brain) |
1 dataset with 18 patients | Intensity-based | Medical image synthesis |
| J. P. Fortin et al. (2018) [28] | MRI (brain) |
11 scanners | Feature-based | Cortical thickness harmonization |
| J. C. Beer et al. (2019) [29] | Structural MRI (brain) |
58 sites | Feature-based | Alzheimer prediction |
| C. Ma et al. (2019) [30] | 3D Cardiac MRI | 20 scans (10 training and 10 test) | Intensity-based | Image segmentation |
| D. Tian et al. (2022) [31] | T1 MRI (brain) |
12 centers | Intensity-based | Gray matter analysis |
| M. Shah et al. (2011) [32] | T1w, T2w, and PDw (brain) |
10 scanners | Intensity-based | Multiple sclerosis lesion identification |
| Liu et al. (2018) [33] | T2w (brain-glioma) |
2 cohorts | Intensity-based | Prediction of progression-free survival in lower-grade gliomas |
| F. Lucia et al. (2019 [34]) | T1, T2 DWI (cervical cancer) |
3 centers | Feature-based | Cervical cancer prognosis |
| Peeken et al. (2019) [35] | Contrast-enhanced T1-weighted fat saturated (T1FSGd), fat-saturated T2-weighted (T2FS) (sarcoma) |
2 cohorts | Intensity/ feature-based | Classification of low and high grade soft tissue sarcoma |
| Liu et al. (2019) [36] | T1w,T2w-fl1air (brain) |
4 centers | Feature-based | Prediction of the individualized treatment response in children with cerebral palsy |
| C. Hognon et al. (2019) [37] | T1, T1c, T2, FLAIR (glioblastoma) |
3 centers | Intensity-based | Image segmentation |
| R. Da-Ano et al. (2020) [38] | Post-injection gadolinium contrast-enhanced MRI (GADO), T2-weighted MRI (T2) and apparent diffusion coefficients (ADC) maps from diffusion-weighted MRI (cervical cancer) |
3 centers | Feature-based | Prediction and treatment adaptation |
| D. Moyer et al. (2020) [39] | Diffusion MRI (brain) |
15 patients from 2 scanners | Intensity-based | White Matter analysis |
| G. Modanwal et al. (2020) [40] | DCE-MRI (breast cancer) |
124 patients from 2 scanners | Intensity-based | Evaluation |
| J. Zhong et al. (2020) [41] | Neonatal DTI-MRI (brain) |
84 neonates data from 2 sites | Deep Learning | Harmonize neonatal data |
| K. Armanious et al. (2020) [42] | T1 (brain) |
11 patients | Intensity-based | Motion correction |
| Scalco et al. (2020) [43] | T1w, T2w (prostate) |
3 different organs of interest | Intensity-based | Reproducibility estimation |
| R. Da-Ano et al. (2021) [44] | Post-injection gadolinium contrast-enhanced MRI (GADO), T2-weighted MRI (T2) and apparent diffusion coefficients (ADC) maps from diffusion-weighted MRI (cervical cancer) | 3 centers | Feature-based | Prediction |
| Saint Martin et al. (2021) [45] | T1, T2, T1-DCE (breast) |
2 phantoms (2 scanners and 3 dual breast coils) |
Both | Lesion classification |
| N.K. Dinsdale et al. (2021) [46] | T1w (neuroimages) |
3 dataset centers | Feature-based | Age prediction and segmentation |
| F. Orlhac et al. (2021) [47] | T1, FLAIR and contrast-enhanced T1-weighted (CE-T1w) images and T2w (brain/prostate) |
2 centers | Feature-based | Impact of harmonization to distinguish between Gleason grades |