A federated framework of brain image processing using the Global Alzheimer's Association Interactive Network

Abstract

Abstract Background The Global Alzheimer's Association Interactive Network (GAAIN) is a unique federated framework that allows users to explore clinical and biomarker data from 57 Alzheimer’s disease data repositories around the world. Integrating brain image processing pipelines with GAAIN data is an important next step that will allow users to derive imaging metrics based on their own analyses. Method To that end, we present a federated image processing framework that expands the current capabilities of GAAIN. Within this framework, GAAIN users can initiate image processing pipelines remotely on the data partners’ imaging data prior to requesting the data. Within the same framework, GAAIN users can also explore the results derived from each imaging pipeline and perform basic analyses. Finally, results from each analysis can be imported back to GAAIN and analyzed alongside existing GAAIN variables. Result We present a case study by initiating a volumetric analysis directly from GAAIN. Conclusion This framework provides a solution for imaging data owners to house and protect their imaging data in compliance with the federated character of GAAIN, and, at the same time, for users to analyze imaging data relevant to their GAAIN cohort of interest. We believe that deploying brain image processing pipelines on existing GAAIN data repositories will be pivotal in terms of using big data to propel basic research and clinical trial design in Alzheimer’s disease.

Ähnliche Arbeiten