Profile-tracking analysis (PTA) estimates nonlinear trajectories, trends or profiles through smoothing splines. Currently the program PTA only works through a command-line scripting mode. It shares the same modeling mechanism as the program 3dMSS. The differences are: 1) 3dMSS is usually used for 3D or 1D data (e.g., surface data in .niml.dset, 20 or more ROIs) while PTA is designed for ROI (less than 20) or behavior data at the subject or population level; 2) the output from PTA can be easily used for customized demonstration of estimated trajectories (with uncertainty bands).
The usage and examples can be found in the help of the program:
PTA -help | less
The underlying theory is covered in the following paper:
Chen et al. (2020). Beyond linearity: Capturing nonlinear relationships in neuroimaging. https://doi.org/10.1101/2020.11.01.363838