Abstract:
|
Imaging-based brain connectivity measures have become an important tool for investigating the pathophysiology, progression and treatment response of psychiatric disorders. In this work, we investigate the predictability of the functional connectivity in resting-state fMRI (rs-fMRI). Specifically, we propose a prediction method based on Bayesian hierarchical model that uses individual's earlier scans, coupled with relevant baseline characteristics, to predict the individual's future functional connectivity. The proposed prediction method could provide a useful tool to predict the changes in individual patient's brain connectivity with the progression of the disease. It can also be used to predict a patient's brain connectivity after a specified treatment regimen which could potentially help guide individualized treatment plan. Another utility of the proposed method is that it could be applied to test-retest imaging data to develop a more reliable estimator for individual functional connectivity. We illustrate the application of the methods with the longitudinal rs-fMRI from ADNI2 study and also with the test-retest rs-fMRI from Kirby21 study.
|