rsHRF: A toolbox for resting-state HRF estimation and deconvolution
Guo-Rong Wu,
Nigel Colenbier,
Sofie Van Den Bossche,
Kenzo Clauw,
Amogh Johri,
Madhur Tandon,
Daniele Marinazzo
Affiliations
Guo-Rong Wu
Key Laboratory of Cognition and Personality, Faculty of Psychology, Southwest University, Chongqing 400715, China; Department of Data Analysis, Faculty of Psychology and Educational Sciences, Ghent University, Ghent 9000, Belgium; Corresponding author at: Key Laboratory of Cognition and Personality, Faculty of Psychology, Southwest University, Chongqing 400715, China.
Nigel Colenbier
Department of Data Analysis, Faculty of Psychology and Educational Sciences, Ghent University, Ghent 9000, Belgium; Research Center for Motor Control and Neuroplasticity, KU Leuven, Leuven 3001, Belgium; Brain Imaging and Neural Dynamics Research Group, IRCCS San Camillo Hospital, Venice 30126, Italy
Sofie Van Den Bossche
Department of Data Analysis, Faculty of Psychology and Educational Sciences, Ghent University, Ghent 9000, Belgium
Kenzo Clauw
Department of Data Analysis, Faculty of Psychology and Educational Sciences, Ghent University, Ghent 9000, Belgium
Amogh Johri
International Institute of Information Technology, Bangalore 560100, India
Madhur Tandon
Indraprastha Institute of Information Technology, Delhi 110020, India
Daniele Marinazzo
Department of Data Analysis, Faculty of Psychology and Educational Sciences, Ghent University, Ghent 9000, Belgium; Brain Imaging and Neural Dynamics Research Group, IRCCS San Camillo Hospital, Venice 30126, Italy
The hemodynamic response function (HRF) greatly influences the intra- and inter-subject variability of brain activation and connectivity, and might confound the estimation of temporal precedence in connectivity analyses, making its estimation necessary for a correct interpretation of neuroimaging studies. Additionally, the HRF shape itself is a useful local measure. However, most algorithms for HRF estimation are specific for task-related fMRI data, and only a few can be directly applied to resting-state protocols. Here we introduce rsHRF, a Matlab and Python toolbox that implements HRF estimation and deconvolution from the resting-state BOLD signal. We first provide an overview of the main algorithm, practical implementations, and then demonstrate the feasibility and usefulness of rsHRF by validation experiments with a publicly available resting-state fMRI dataset. We also provide tools for statistical analyses and visualization. We believe that this toolbox may significantly contribute to a better analysis and understanding of the components and variability of BOLD signals.
