search

james-cole / brainageR

Software for generating a brain-predicted age value, using Gaussian Processes regression, implemented in R
GNU Lesser General Public License v3.0
76 stars 27 forks source link

RH Age #8

Open rogiedodgie opened 2 years ago

rogiedodgie commented 2 years ago

Would it be possible to train the models on RH and LH separately. This would be great for those looking at unilateral stroke. The ability to chart the health of the intact hemisphere over time could really help us get a grasp of how stroke in one hemisphere affects the other. Best, Roger

james-cole commented 2 years ago

Hi - good question!

Currently, the brainageR model is whole-brain only, so in short the answer is no. The long answer is yes, it would be possible to re-train the model using only one hemisphere or the other. That's not something we're planning for this software, but we do have alternatives, including the the local brain-age approach: https://github.com/SebastianPopescu/U-NET-for-LocalBrainAge-prediction

rogiedodgie commented 2 years ago

Got it, a follow up question. If we average the brainage of all RH ROIs (using code form the cited repo), would that be equivalent to the RH brain age? Thanks, Roger

james-cole commented 2 years ago

Another good question - actually no, those ROIs probably wouldn't cover the entirety of a single hemisphere. It would be better to split a whole brain mask down the middle use that as a hemisphere mask

rogiedodgie commented 2 years ago

Thank you James, I’ll give it a go! Roger

On Fri, Nov 11, 2022 at 4:28 AM James Cole @.***> wrote:

Another good question - actually no, those ROIs probably wouldn't cover the entirety of a single hemisphere. It would be better to split a whole brain mask down the middle use that as a hemisphere mask

— Reply to this email directly, view it on GitHub https://github.com/james-cole/brainageR/issues/8#issuecomment-1311447888, or unsubscribe https://github.com/notifications/unsubscribe-auth/AGIXRRM5S7KQZ5UKKDY3QF3WHYGS3ANCNFSM6AAAAAAR3YL3IU . You are receiving this because you authored the thread.Message ID: @.***>