Hello,

In the context of both cardiac and knee challenges, there are multiple acceleration rates available. For cardiac scans, there are 4, 8, and 10-fold accelerated scans, and for knee scans, there are 4 and 8-fold accelerated scans. I'm seeking clarification on the training and testing procedure for the network:

Was the network trained exclusively with 4-fold accelerated scans and then tested with 4, 8, and 10-fold accelerated scans?

Or, was the network trained with 4-fold accelerated scans and tested only with 4-fold accelerated test scans?

Additionally, I'd like to know if the same approach extends to other acceleration rates.

Thank you,

Thanks for your amazing work, which has inspired me a lot！
I am currently trying to apply PromptMR to the calgary-campinas dataset and examine its effects. However, I am facing some difficulties in processing the data and adapting it to the format required by the model. If possible, could you please share the code for training PromptMR on this dataset?
I appreciate it very much ！

When I run the pretrained file, an error occurs:
Traceback (most recent call last):
File "run_pretrained_promptmr_fastmri_knee_inference.py", line 182, in
torch.device(args.device),
File "run_pretrained_promptmr_fastmri_knee_inference.py", line 121, in run_inference
model.load_state_dict(torch.load(state_dict_file))
File "/home/zhengkaiyi/anaconda3/envs/zky/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1672, in load_state_dict
self.class.name, "\n\t".join(error_msgs)))
RuntimeError: Error(s) in loading state_dict for PromptMR:
Is it because I ran the files in the wrong order, or is there another reason? Could you help me solve it?

Hello,

What spatial dimensions are you using for training on the CMRxRecon dataset? I didn't find any cropping or padding in the data loaders for it.

Is it purely through asymmetric padding in the Unit Layers for the full-resolution scans? [144x512 or 116x382 for the T1/T2 task]

Hello,

I am interested in using non-Cartesian trajectories (radial, spiral) in imaging and would like to know if PromptMR can support these types of trajectories. This feature could add significant value to our current workflows.

Best,

Hi,
Thanks for sharing the code of this interesting work.

I am trying to run the training on the fastMRI dataset and I got CUDA out of memory issue even with batch size=1.
My GPU is NVIDIA A5000, which has 24G memory.

Could you please tell me how much GPU memory is required to train with batchsize=1?

BTW, I noticed that the memory is gradually increasing for each iteration (batch).
Is that normal? Maybe this is somehow related to the code itself and I didn't notice.

Many thanks! looking forward to your reply.

Hello,

Is it possible to share 4x acceleration knee checkpoint ? I want to compare it with ensemble methods and prospective other tasks.

Thanks,

Hello,

Thank you very much for your effort and congratulations for this amazing achievement.

I have two questions if you allow me.
1) Regrading the validation set, the challenge submission leaderboard is closed now. May I ask if is there a way to reproduce
the results you reported in the paper (I did the training already)?
2) May I know the system you used for the training please?

Thank you very much for your time.