Table 1.
Benchmark datasets for federated learning in medical image analysis.
| Reference | Task | Dataset | Modality | Learning model |
|---|---|---|---|---|
| Brain | ||||
| Peng et al. (2022) [81] | ASD, AD classification | ABIDE [103], ADNI [101] | fMRI | GCN |
| Gürler et al. (2022) [120] | Brain connectivity prediction | OASIS [121] | MRI | GNN |
| Islam et al. (2022) [122] | Brain tumor classification | UK Data Service [123] | MRI | CNN |
| Dinsdale et al. (2022) [49] | Age prediction | ABIDE [103] | MRI | CNN (VGG) |
| Jiang et al. (2022) [124] | Intracranial hemorrhage diagnosis | RSNA [105] | CT | CNN (DenseNet) |
| Stripelis et al. (2021) [36] | Brain age prediction | UK Biobank [106] | MRI | CNN |
| Liu et al. (2021) [125] | Intracranial hemorrhage diagnosis | RSNA [105] | CT | CNN (DenseNet) |
| Fan et al. (2021) [85] | ASD classification | ABIDE [103] | MRI | CNN |
| Li et al. (2020)[48] | ASD classification | ABIDE [103] | fMRI | MLP |
| Sheller et al. (2019) [126] | Brain tumor segmentation | BraTS [104] | MRI | U-Net |
| Li et al. (2019) [90] | Brain tumor segmentation | BraTS [104] | MRI | CNN |
| Chest | ||||
| Hatamizadeh et al. (2023) [94] | Image generation (attack) | COVID CXR [109] | Chest X-ray | CNN (ResNet) |
| ChestX-ray14 [108] | ||||
| Yan et al. (2023) [33] | Classification | COVID-FL [33] | Chest X-ray | Transformer |
| Alkhunaizi et al. (2022) [32] | Classification | CheXpert [107] | Chest X-ray | CNN |
| Dong et al. (2022) [127] | Classification | ChestX-ray14 [108] | Chest X-ray | CNN |
| Chakravarty et al. (2021) [43] | Classification | CheXpert [107] | Chest X-ray | CNN, GNN |
| Lung | ||||
| Yang et al. (2021) [35] | COVID-19 diagnosis | COVIDx [110] | Chest X-ray | CNN |
| Feki et al. (2021) [128] | COVID-19 diagnosis | Local dataset | Chest X-ray | CNN |
| Kumar et al. (2021) [76] | COVID-19 diagnosis | COVID-19 CXR [109] | Chest X-ray | CNN |
| Dong et al. (2021) [61] | COVID-19 diagnosis | COVID-19 CXR [109] | Chest X-ray | CNN |
| Yang et al. (2021) [72] | Segmentation | Local dataset | CT | CNN |
| Heart | ||||
| Linardos et al. (2022) [129] | Cardiac diagnosis | ACDC [111], M&M [112] | MRI | CNN |
| Qi et al. (2022) [124] | Cardiac segmentation | M&M [112], Emidec [130] | MRI | U-Net |
| Li et al. (2021) [131] | Cardiac image synthesis | Local dataset | CT | GAN |
| Wu et al. (2021) [59] | Cardiac segmentation | ACDC [111] | MRI | U-Net |
| Breast | ||||
| Agbley et al. (2023) [132] | Breast tumor classification | BreakHis [133] | Pathology | CNN |
| Wicaksana et al. (2022) [134] | Breast tumor segmentation | BUS [135], BUSIS [136], UDIAT [137] | Ultrasound | U-Net |
| Skin | ||||
| Yan et al. (2023) [33] | Skin lesion classification | ISIC [114] | Dermoscopy | Transformer |
| Wicaksana et al. (2022) [39] | Skin lesion classification | HAM10000 [113] | Dermoscopy | CNN |
| Alkhunaizi et al. (2022) [32] | Skin lesion classification | HAM10000 [113] | Dermoscopy | CNN |
| Jiang et al. (2022) [124] | Skin lesion classification | HAM10000 [113] | Dermoscopy | CNN (DenseNet) |
| Liu et al. (2021) [125] | Skin lesion classification | HAM10000 [113] | Dermoscopy | CNN (DenseNet) |
| Bdair et al. (2021) [138] | Skin lesion classification | HAM10000 [113] | Dermoscopy | CNN (EfficientNet) |
| ISIC19 [139] | ||||
| Derm7pt [140] | ||||
| PAD-UFES [141] | ||||
| Chen et al. (2021) [142] | Skin lesion classification | HAM10000 [113] | Dermoscopy | CNN (VGG) |
| ISIC17 [143] | ||||
| Eye | ||||
| Yan et al. (2023) [33] | Diabetic classification | Retina [144] | Color retinal image | Transformer |
| Qiu et al. (2023) [145] | Fundus segmentation | RIM-ONE [146] | Color retinal image | CNN (MobileNet) |
| Wang et al. (2023) [147] | Fundus segmentation | RIF [148] | Color retinal image | Transformer |
| Qu et al. (2022) [52] | Diabetic classification | Retina [144] | Color retinal image | VAE, CNN |
| Abdomen | ||||
| Zhu et al. (2023) [149] | Prostate segmentation | PROMISE12 [115] | MRI | U-Net |
| NCI-ISBI 2013 [150] | ||||
| Qiu et al. (2023) [145] | Prostate segmentation | PROMISE12 [115] | MRI | CNN (MobileNet) |
| Xu et al. (2023) [151] | Tumor segmentation | LiTS [152] | CT | U-Net |
| Wicaksana et al. (2022) [39] | Cancer classification | ProstateX [153] | MRI | CNN |
| Luo et al. (2022) [86] | Cancer classification | OrganMNIST [119] | CT | CNN |
| Liu et al. (2022) [154] | Polyp detection | GLRC [155] | Colonoscopy | CNN |
| Yan et al. (2021) [53] | Cancer classification | ProstateX [153] | MRI | GAN |
| Roth et al. (2021) [156] | Prostate segmentation | MSD-Prostate [157] | MRI | U-Net |
| PROMISE12 [115] | ||||
| ProstateX [153] | ||||
| NCI-ISBI 2013 [150] | ||||
| Histology | ||||
| Hosseini et al. (2023) [84] | Cancer classification | TCGA [116] | Pathology | CNN (DenseNet) |
| du Terrail et al. (2023) [158] | Cancer classification | Local dataset | Pathology | CNN |
| Lu et al. (2022) [73] | Cancer classification | TCGA [116] | Pathology | CNN |
| Adnan et al. (2022) [159] | Cancer classification | TCGA [116] | Pathology | CNN (DenseNet) |
| Luo et al. (2022) [86] | Cancer classification | PathMNIST [119] | Pathology | CNN |
| Wagner et al. (2022) [47] | Image harmonization | PESO [160] | Pathology | GAN |
| Ke et al. (2021) [46] | Image harmonization | TCGA [116] | Pathology | GAN |
| Others | ||||
| Feng et al. (2022) [41] | MRI reconstruction | fastMRI [117] | MRI | U-Net |
| BraTS [104] | ||||
| Elmas et al. (2022) [161] | MRI reconstruction | fastMRI, BraTS, IXI | MRI | GAN |
| Guo et al. (2021) [51] | MRI reconstruction | fastMRI, BraTS, IXI | MRI | U-Net |