Trouble understanding and accessing PICMUS dataset

[dewangbhanushali-25]

Hi
I’m an undergraduate student interested in medical signal processing. I saw that your project uses the PICMUS dataset, and I’m trying to learn how to work with it.

Right now, I’m not able to figure out how to access and understand the dataset. It’s confusing for me to see how the files are organized and how to use them.

Could you please help explain:

How to get started with the PICMUS data?

Which parts are publicly available and usable for research?

Any examples or documentation you recommend?

I tried reaching out on LinkedIn but didn’t get a reply. Is GitHub the best place to ask?

Thank you very much for your help!

Best regards,

[tristan-deep]

Hi @dewangbhanushali-25,

How to get started with the PICMUS data?

An example of using the PICMUS dataset with zea is given in this notebook.

Which parts are publicly available and usable for research?

The entire zea repository is publicly available. Obviously if you use a dataset or a model provided here, please ensure appropriate attribution to the authors (for instance cite PICMUS in this case). You can also cite this repository if you find it helpful. See details here.

Any examples or documentation you recommend?

I can recommend reading the docs, and the example page specifically. All examples should run out of the box in colab.

Hope that helps!

[dewangbhanushali-25]

I appreciate your time and assistance. Thank you sir !

[dewangbhanushali-25]

Hello everyone,

I've recently downloaded the PICMUS dataset and noticed that it contains very few raw data files. I'm curious about how the community was able to reconstruct images and develop robust models given this apparent limitation. Specifically:

Were there any data augmentation techniques or external resources used to supplement the dataset?

Did you face challenges in training models due to the small size, and how did you overcome them?

I apologize for any inconvenience my questions may cause and appreciate any guidance or experiences you can share regarding working with the PICMUS dataset. Thank you for your understanding and support!