Model Overview
UNet for 12x accelerated MRI Reconstruction on the StanfordKnees2019 dataset.
ATOMMIC: Training
To train, fine-tune, or test the model you will need to install
ATOMMIC . We recommend you install it after you've installed latest Pytorch version.
pip install atommic['all']How to Use this Model
The model is available for use in ATOMMIC, and can be used as a pre-trained checkpoint for inference or for fine-tuning on another dataset.
Corresponding configuration YAML files can be found
here .
Automatically instantiate the model
1 pretrained: true
2 checkpoint: https://huggingface.co/wdika/REC_UNet_StanfordKnees2019_gaussian2d_12x_AutoEstimationCSM/blob/main/REC_UNet_StanfordKnees2019_gaussian2d_12x_AutoEstimationCSM.atommic
3 mode: test
Usage
You need to download the Stanford Knees 2019 dataset to effectively use this model. Check the
StanfordKnees2019 page for more information.
Model Architecture
1 model:
2 model_name: UNet
3 channels: 64
4 pooling_layers: 4
5 in_channels: 2
6 out_channels: 2
7 padding_size: 11
8 dropout: 0.0
9 normalize: true
10 norm_groups: 2
11 dimensionality: 2
12 reconstruction_loss:
13 wasserstein: 1.0
Training
1 optim:
2 name: adamw
3 lr: 1e-4
4 betas:
5 - 0.9
6 - 0.999
7 weight_decay: 0.0
8 sched:
9 name: InverseSquareRootAnnealing
10 min_lr: 0.0
11 last_epoch: -1
12 warmup_ratio: 0.1
13
14 trainer:
15 strategy: ddp_find_unused_parameters_false
16 accelerator: gpu
17 devices: 1
18 num_nodes: 1
19 max_epochs: 20
20 precision: 16-mixed
21 enable_checkpointing: false
22 logger: false
23 log_every_n_steps: 50
24 check_val_every_n_epoch: -1
25 max_steps: -1
Performance
To compute the targets using the raw k-space and the chosen coil combination method, accompanied with the chosen coil sensitivity maps estimation method, you can use
targets configuration files.
Evaluation can be performed using the
evaluation script for the reconstruction task, with --evaluation_type per_slice.
Results
Evaluation against SENSE targets
12x: MSE = 0.001251 +/- 0.005686 NMSE = 0.04254 +/- 0.09148 PSNR = 31.4 +/- 6.554 SSIM = 0.7705 +/- 0.2946
Limitations
This model was trained on the StanfordKnees2019 batch0 using a UNet coil sensitivity maps estimation and Geometric Decomposition Coil-Compressions to 1-coil, and might differ from the results reported on the challenge leaderboard.
References
[2] Epperson K, Rt R, Sawyer AM, et al. Creation of Fully Sampled MR Data Repository for Compressed SENSEing of the Knee. SMRT Conference 2013;2013:1
