Randomized Structural Sparsity based Support Identification with Applications to Locating Activated or Discriminative Brain Areas: A Multi-center Reproducibility Study

TL;DR

This study introduces a robust, stable method for identifying brain regions related to stimuli or diseases using multi-center neuroimaging data, overcoming high dimensionality and inter-center variability.

Contribution

It applies a recently developed stability selection and structural sparsity algorithm to multi-center MRI data, demonstrating improved reproducibility and robustness over existing methods.

Findings

Achieved consistent brain region identification across multiple centers.

Demonstrated robustness and insensitivity to key algorithm parameters.

Identified interpretable biomarkers from functional and structural MRI data.

Abstract

In this paper, we focus on how to locate the relevant or discriminative brain regions related with external stimulus or certain mental decease, which is also called support identification, based on the neuroimaging data. The main difficulty lies in the extremely high dimensional voxel space and relatively few training samples, easily resulting in an unstable brain region discovery (or called feature selection in context of pattern recognition). When the training samples are from different centers and have betweencenter variations, it will be even harder to obtain a reliable and consistent result. Corresponding, we revisit our recently proposed algorithm based on stability selection and structural sparsity. It is applied to the multi-center MRI data analysis for the first time. A consistent and stable result is achieved across different centers despite the between-center data variation while many other state-of-the-art methods such as two sample t-test fail. Moreover, we have empirically showed that the performance of this algorithm is robust and insensitive to several of its key parameters. In addition, the support identification results on both functional MRI and structural MRI are interpretable and can be the potential biomarkers.