search

abrilcf / mednerf

MIT License
139 stars 30 forks source link

CT generation #9

Open dsitnik-mango opened 1 year ago

dsitnik-mango commented 1 year ago

Hi,

Thanks for the code! Is it possible to predict a CT volume instead of different views (DRRs) once the network is trained?

I want to give the trained network a single DRR and predict a volume cube. Is that possible with your model? If not, where should I make the modifications?

Thank you for your help in advance! Dario

ro1406 commented 12 months ago

Hi @dsitnik-mango! Did you manage to figure out if this is possible with the model? If yes, could you guide me on the same? If not, were there any other papers with code that could do this? Thanks!

dsitnik-mango commented 12 months ago

@ro1406 With this model, not yet...let me know if you find a solution

ro1406 commented 12 months ago

@dsitnik-mango Sure. Is there another model you were able to find that generates 3D volumes from a 2D input? If so I'd appreciate it if you could share the repo :)

Zhentao-Liu commented 3 months ago

I also have the same question. But could we do marching cubes with the well-trained generative model? FYI: build a voxel grids, query coordinates density value, and then form a 3D CT volume.