Whole-body metabolic connectivity framework with functional PET.

Understanding how organs communicate metabolically is a growing research priority. Prior research has shown that the brain and organs interact through complex networks. However, the exact mechanisms of these interactions remain unclear. No prior work had resolved how to quantify these interactions systematically. This gap motivated the development of a framework to estimate metabolic connectivity. Existing methods lack the ability to capture dynamic interactions across multiple organs. The nervous and circulatory systems are known to facilitate communication between organs. Yet, how these systems influence each other under normal and pathological conditions is still uncertain.

The aim of this study was to develop a framework for estimating metabolic connectivity between organs and brain regions. The researchers sought to address the lack of standardized methods for analyzing whole-body metabolic interactions. They focused on creating a reproducible workflow using functional PET imaging. This approach could help identify how organ dysfunction affects the entire system. The study aimed to compare different segmentation and connectivity estimation techniques. The researchers wanted to determine which methods provide the most reliable results. They also aimed to explore brain-body interactions in healthy individuals. This work could lead to better diagnostic tools for complex disorders involving metabolic dysregulation.

The study used functional PET scans from 16 healthy subjects. Researchers applied dynamic 18F-FDG PET/CT imaging to capture metabolic activity. They tested multiple segmentation techniques, including manual and automated methods. Spatiotemporal filtering and polynomial fitting were used to estimate connectivity. Covariance matrices were calculated to assess inter-organ relationships. The team compared circular volumes with manual segmentation for accuracy. They evaluated how different methods affect connectivity estimates. The framework was optimized to provide a comprehensive overview of metabolic interactions.

Automated organ delineation showed high agreement with manual segmentation. Polynomial fitting produced similar connectivity results as spatiotemporal filtering. However, covariance matrices at the group level did not match individual results. The strongest brain-body connectivity was observed in the liver and kidneys. These findings suggest that these organs play a central role in metabolic communication. The liver's connectivity with brain regions was particularly notable. Kidney interactions also showed significant metabolic influence. The framework successfully captured both intra-organ and inter-organ connectivity patterns.

The proposed framework provides a novel way to estimate metabolic connectivity in a systemic manner. The results suggest that automated segmentation is reliable and efficient. The study found that liver and kidney connectivity with the brain is strong and consistent. These findings may help in understanding how organ dysfunction affects the whole body. The mismatch between individual and group-level covariance matrices indicates variability in connectivity patterns. This variability could reflect individual metabolic differences or methodological limitations. The framework's ability to capture hierarchical interactions is a key contribution. Future work could explore how these patterns change in pathological states.