A group comparison in fMRI data using a semiparametric model under shape invariance

Arunava Samaddar, Brooke S. Jackson, Christopher J. Helms, Nicole A. Lazar, Jennifer E. McDowell, Cheolwoo Park

Research output: Contribution to journalArticlepeer-review


In the analysis of functional magnetic resonance imaging (fMRI) data, a common type of analysis is to compare differences across scanning sessions. A challenge to direct comparisons of this type is the low signal-to-noise ratio in fMRI data. By using the property that brain signals from a task-related experiment may exhibit a similar pattern in regions of interest across participants, a semiparametric approach under shape invariance to quantify and test the differences in sessions and groups is developed. The common function is estimated with local polynomial regression and the shape invariance model parameters are estimated using evolutionary optimization methods. The efficacy of the semi-parametric approach is demonstrated on a study of brain activation changes across two sessions associated with practice-related cognitive control. The objective of the study is to evaluate neural circuitry supporting a cognitive control task, and associated practice-related changes via acquisition of blood oxygenation level dependent (BOLD) signal collected using fMRI. By using the proposed approach, BOLD signals in multiple regions of interest for control participants and participants with schizophrenia are compared as they perform a cognitive control task (known as the antisaccade task) at two sessions, and the effects of task practice in these groups are quantified.

Original languageEnglish (US)
Article number107361
JournalComputational Statistics and Data Analysis
StatePublished - Mar 2022

All Science Journal Classification (ASJC) codes

  • Statistics and Probability
  • Computational Mathematics
  • Computational Theory and Mathematics
  • Applied Mathematics


Dive into the research topics of 'A group comparison in fMRI data using a semiparametric model under shape invariance'. Together they form a unique fingerprint.

Cite this