You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi everyone! I have recently reworked nitrain (previously torchsample) to better support modern workflows for medical imaging AI using both keras and pytorch.
I've made a ton of new additions. First, I've added samplers which allow you to seamlessly train models on only parts of images - i.e., slices, patches, or blocks - without changing any part of your workflow. Next, I've added better datasets support so that you can intuitively grab image data from folders and other sources. Finally, I'm expanded the trainers to allow users to launch cloud training jobs directly from nitrain.
With that said - any feedback would be welcome! I am particularly interested in hearing about what aspects of medical imaging AI are still difficult even with existing tools.
Feel free to respond here or reach me at nickcullen31 at gmail dot com. Thanks!
The text was updated successfully, but these errors were encountered:
Hi everyone! I have recently reworked nitrain (previously torchsample) to better support modern workflows for medical imaging AI using both keras and pytorch.
I've made a ton of new additions. First, I've added
samplers
which allow you to seamlessly train models on only parts of images - i.e., slices, patches, or blocks - without changing any part of your workflow. Next, I've added betterdatasets
support so that you can intuitively grab image data from folders and other sources. Finally, I'm expanded thetrainers
to allow users to launch cloud training jobs directly from nitrain.With that said - any feedback would be welcome! I am particularly interested in hearing about what aspects of medical imaging AI are still difficult even with existing tools.
Feel free to respond here or reach me at nickcullen31 at gmail dot com. Thanks!
The text was updated successfully, but these errors were encountered: