search

ncoudray / DeepPATH

Classification of Lung cancer slide images using deep-learning
492 stars 213 forks source link

Influence on parameters #20

Closed PradeepMoturi closed 5 years ago

PradeepMoturi commented 5 years ago

How does tiles from LUAD/LUSC slide without tumour cells influences the parameters?

ncoudray commented 5 years ago

Which parameters? Can you please clarify your question?

PradeepMoturi commented 5 years ago

How do they influence inception network parameters (values in kernal matrices)? I mean, the tiles with no tumor region in LUAD, LUSC slide shouldn't be influencing the weights.

ncoudray commented 5 years ago

If you don't want them to influence the weights, then you need to to remove them - the easiest way would be to do a pre-selection of the ROI using Aperio, save the coordinates in a xml file which can then be used an input of the "tiling" script.

ncoudray commented 5 years ago

If you don't want them to influence the weights, then you need to to remove them - the easiest way would be to do a pre-selection of the ROI using Aperio, save the coordinates in a xml file which can then be used an input of the "tiling" script.

bcli4d commented 5 years ago

Can you explain "using Aperio". Is that some (free?) Leica software? If yes, is it the software that is doing the segmentation or must that be done by the user?

ncoudray commented 5 years ago

Hi,

See https://www.leicabiosystems.com/digital-pathology/manage/aperio-imagescope/ for Aperio imagescope software.