nilearn-fmri

To run a first-level GLM on task-based fMRI data, provide your NIfTI files along with events and confounds. The analysis fits a nilearn GLM to compute z or t contrast maps for specified conditions.

Yes, you can compute resting-state functional connectivity by extracting ROI time series from atlases like Schaefer or AAL. It applies bandpass filtering and detrending before generating connectivity matrices for group comparisons.

You can perform MVPA decoding by extracting voxel-level or connectivity-derived features and training a support vector machine (SVM) classifier via scikit-learn to identify patterns distinguishing experimental conditions.

Yes, CanICA works for spatial ICA decomposition by extracting independent component maps from fMRI datasets. It identifies spatially independent brain networks directly from preprocessed NIfTI inputs.

You need nilearn, nibabel, numpy, scipy, pandas, scikit-learn, matplotlib, and joblib installed. These scientific Python dependencies support NIfTI loading, statistical modeling, and visualization.

Yes, you can visualize group-level z-maps after running a second-level GLM. The workflow computes one-sample group analyses from first-level contrasts and generates visual outputs using matplotlib.