Training models to detect multiple classes

I have requirement where I need to detect and segment different sub disease parts. The catch is, one CT scan data contains only sub type of disease. If the sub-diseases exists in all images, its fairly possible and is straight forward.
In my scenario, do you think one model can able to detect each of subtypes or do I need more models?


You can do this in multiple ways:

  • Train multiple models each looking for one disease in a specific organ
    or do a pipeline of models as
  • model to segment all organs or sub set of organs( just normal anatomy)
  • another model take in 1 organ and segment abnormalities
  • another model to classify the abnormalities to different disease levels

Hope that helps

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.