We use high resolution structural brain MRIs and a murine polymicrobial sepsis LD50 model to demonstrate that pre-infection variability in brain structure can reliably predict infection outcome. Specifically, mice fated to survive exhibited greater cortical volume and thickness compared to those who succumbed. Our work reveals a readily measurable trait that can predict a murine LD50 polymicrobial sepsis outcome
To predict a murine LD50 polymicrobial sepsis, we process our MRI images using DeepBrainIPP. First, we update DeepBrainIPP model with ourdataset to perform skull Stripping. Then we perform images registration and morphological analysis. Finally, we build predictive models with measured morphologies.
We shared the complete data on morphological measurements as well as a python scripts (see "scripts folder") with step by step description on how models were built. The details can be found in our manuscripts (under review).
Please feel free to contact ([email protected], [email protected]) if you have questions