Journal: Nature protocols
Article Title: Toward a unified framework for interpreting machine-learning models in neuroimaging
doi: 10.1038/s41596-019-0289-5
Figure Lengend Snippet: Key functions in the CanlabCore MATLAB toolbox used in the protocol
Article Snippet: Table 3 | Function Description Used in... filenames Lists file names that match a specific pattern in a directory Step 1A; Step 8 fmri_data Loads images into the object or creates an empty fmri_data object Step 1A; Step 5; Step 7C; Step 8; Step 11A apply_mask Masks an image with a defined mask or calculates pattern expression values Step 1A; Step 9 predict Trains and evaluates predictive models and performs cross-validation and bootstrap tests Step 1A; Step 2; Step 6; Step 7A roc_plot Calculates accuracy, sensitivity and specificity and visualizes the ROC curve Step 3; Step 7C; Step 10; Step 12 threshold Applies a statistically valid threshold to the images with statistical test results Step 7A orthviews Displays the image data stored in CANlab tools Step 7A; Step 7B write Writes data stored in the fmri_data or statistic_image objects into a NIfTI (.nii) or Analyze (.img) file Step 7A; Step 7B svm_rfe Performs recursive feature elimination with support vector machines Step 7B canlab_pattern_similarity Calculates similarity between each column in a data matrix and a vector of pattern weights Step 11A Open in a separate window Key functions in the CanlabCore MATLAB toolbox used in the protocol.
Techniques: Expressing, Plasmid Preparation