Applying Clinica Machine Learning Classification Results to Clinical Data Analysis

[mtmmu (mtmmu)]

Dear Clinica Project Team,

Greetings!

I am a researcher who has been using Clinica for the classification of Alzheimer's Disease (AD) and Cognitively Normal (CN) subjects. Following the tutorial on your website, I have successfully run the SVM-based machine learning classification pipeline and output the results to a folder.

I am now looking to further apply these classification results to analyze clinical data in the real world. Specifically, I am interested in understanding how to apply these classification models to a new set of clinical data and how to interpret and utilize the output results to support clinical decision-making.

Could you provide some guidance or suggestions? For instance, how to integrate a new dataset into the existing classification framework, and how to understand and utilize the output classification results?

Thank you very much for your assistance!

Sincerely,

mtmmu88

[ninon....@gmail.com]

Dear mtmmu88,

To best answer you we would need a few more details.

First, could you tell us more about your data? Are these imaging data together with clinical data (e.g. demographics, diagnosis)? Are they organised following the BIDS specification? Are they preprocessed already?

Then, what is the task you have in mind exactly? When you say that you want "apply these classification results", do you mean reusing a classifier trained to differentiate AD patients from CN subjects or do you mean re-training a classifier on another task?

Best wishes,

Ninon