ADHD Classification on Emotion and Development Branch Phenotyping and DTI (5-fold cross-validated)
71.9AccuracyFMCIB + MLP + LoRA
Evaluation results
| Method | Links | ||
|---|---|---|---|
| FMCIB + MLP + LoRACategory=FM, Pre-trained Modality=CT, Trainable Params (M)=1.64, FLOPs (T)=0.41, Optimization=ACC-optimized2025.11 | 71.9 | 63 | |
| FMCIB + MLP + LoRACategory=FM, Pre-trained Modality=CT, Trainable Params (M)=1.64, FLOPs (T)=0.41, Optimization=AUC-optimized2025.11 | 63 | 71.6 | |
| 3D U-Net (Models Genesis)Category=DL, Pre-trained Modality=CT, Trainable Params (M)=19.34 (11.79 ×), FLOPs (T)=1.962025.11 | 60 | 55.7 | |
| 3D ResNet-50 (MedicalNet)Category=DL, Pre-trained Modality=CT + MRI, Trainable Params (M)=46.18 (28.16 ×), FLOPs (T)=0.372025.11 | 60 | 57.9 | |
| Decision TreeCategory=ML, Pre-trained Modality=From Scratch2025.11 | 59.3 | 58.5 | |
| 3D ResNet-50Category=DL, Pre-trained Modality=From Scratch, Trainable Params (M)=46.18 (28.16 ×), FLOPs (T)=0.092025.11 | 59.3 | 71 | |
| 3D U-NetCategory=DL, Pre-trained Modality=From Scratch, Trainable Params (M)=19.07 (11.63 ×), FLOPs (T)=0.142025.11 | 57.4 | 67.3 | |
| FMCIB + MLPCategory=FM, Pre-trained Modality=CT, Trainable Params (M)=185.57 (113.15 ×), FLOPs (T)=0.412025.11 | 55.6 | 71.1 | |
| Random ForestCategory=ML, Pre-trained Modality=From Scratch2025.11 | 52.6 | 45.9 | |
| XGBoostCategory=ML, Pre-trained Modality=From Scratch2025.11 | 51.1 | 50.5 | |
| Logistic RegressionCategory=ML, Pre-trained Modality=From Scratch2025.11 | 46.7 | 49 | |
| Support Vector MachineCategory=ML, Pre-trained Modality=From Scratch2025.11 | 43.7 | 48.1 |