01. Sulcal imprints extraction

This is an internal component, see Disco pipeline for general information.

Description

This process extracts the sulcal imprints from the Labelled Cortical Folds Graph of each subject.

Interesting parameters for fine-tuning this step include:

• div_pt_cloud_H, div_pt_cloud_B, allow to modulate the number of points in the imprints. Higher values give less detailed imprints, while smaller values give more detailed imprints (see below for more details).
• max_BL (must be increased if numerous sulci are added to the model, see below).
• ldk_type (see below)

In addition to sulcal imprints, this process creates a text file to summarize the attributes of all studied subjects (subject_hierarchy). This file is notably used by BrainVISA for autocompleting the list of subjects.

This process has hidden outputs, which are not shown at the level of the BrainVISA process. The imprint_workspace.mat files serve as inputs to the next step (02. Template building). Here is a full list:

DISCO_{Study}
└── deformed_output
    └── {subject}-{acquisition}-{analysis}-folds-{graph_version}
        └── {sulci_recognition_session}_auto / {sulci_recognition_session}_manual
            └── mesh
                └── imprint_workspace.mat

Parameters

anatomical_constraints: Choice ( input )

Lgraph: ListOf( Labelled Cortical folds graph ) ( input )

Rgraph: ListOf( Labelled Cortical folds graph ) ( input )

t1mri_nobias: ListOf( T1 MRI Bias Corrected ) ( input )

Lbundles_Volume: ListOf( Bundles Volume ) ( optional, input )

Rbundles_Volume: ListOf( Bundles Volume ) ( optional, input )

study_dir: DISCO Study Dir ( input )

sulci_to_be_studied: Sulci To Be Studied DISCO ( input )

bundles_to_be_studied: Bundles To Be Studied DISCO ( optional, input )

label_translation: Label translation ( input )

sulci_label_translation_siGraph: Sulci Label Translation siGraph DISCO ( input )

left_imprint_mesh: ListOf( Sulcal Imprint Full Mesh ) ( output )

Left sulcal imprints of all subjects in Talairach space.

left_imprint_texture: ListOf( Sulcal Imprint Full Texture ) ( output )

left_imprint_bottom: ListOf( Sulcal Imprint Full Bottom ) ( output )

left_imprint_hull_junction: ListOf( Sulcal Imprint Full Hull Junction ) ( output )

right_imprint_mesh: ListOf( Sulcal Imprint Full Mesh ) ( output )

Right sulcal imprints of all subjects in Talairach space.

right_imprint_texture: ListOf( Sulcal Imprint Full Texture ) ( output )

right_imprint_bottom: ListOf( Sulcal Imprint Full Bottom ) ( output )

right_imprint_hull_junction: ListOf( Sulcal Imprint Full Hull Junction ) ( output )

Talairach_transform: ListOf( Transform Graph to Talairach-AC/PC-Anatomist ) ( output )

div_pt_cloud_H: Float ( input )

Simplification factor for the Hull Junction imprints. Higher values give less detailed imprints, while smaller values give more detailed imprints.

In detail: the imprint is initialized to the set of N points that are contained in the Hull Junction bucket. This point cloud is then simplified to K centroids using a K-means algorithm, where K = floor(N / div_pt_cloud_H) + 1.

Please note that sulcal graphs contain a thick representation of the Hull Junction (typically more than 1 voxel thickness), which means that a higher value of div_pt_cloud_H is needed to obtain a similar imprint resolution compared to div_pt_cloud_B.

div_pt_cloud_B: Float ( input )

Simplification factor for the Bottom imprints. Higher values give less detailed imprints, while smaller values give more detailed imprints.

In detail: the imprint is initialized to the set of N points that are contained in the Bottom bucket. This point cloud is then simplified to K centroids using a K-means algorithm, where K = floor(N / div_pt_cloud_B) + 1.

Please note that sulcal graphs contain a very thin representation of the Bottom (1 voxel thickness), which means that a lower value of div_pt_cloud_B is needed to obtain a similar imprint resolution compared to div_pt_cloud_H.

max_BL: Integer ( input )

Arbitrary value to differenciate hull junction and bottom sulcal imprints. Increase this value if numerous sulci are added to the model.

ldk_type: Choice ( input )

To base the registration only on hull junctions ('H'), only on bottoms ('B'), or both ('B+H'). For DISCO based on bundles only, choose 'H'.

best_way_criteria: Choice ( input )

subjects_hierarchy: Save Subjects Hierarchy DISCO ( output )

Text file to summarize the hierarchy of all studied subjects; used in other steps.

Technical information

Toolbox : Disco

User level : 1

Identifier : sulcalImprintGroup

File name : brainvisa/toolboxes/disco/processes/DISCO_components/sulcalImprintGroup.py

Supported file formats :

Lgraph :

Graph and data, Graph and data

Rgraph :

Graph and data, Graph and data

t1mri_nobias :

NIFTI-1 image, NIFTI-1 image

Lbundles_Volume :

NIFTI-1 image, NIFTI-1 image

Rbundles_Volume :

NIFTI-1 image, NIFTI-1 image

study_dir :

Directory, Directory

sulci_to_be_studied :

Text file, Text file

bundles_to_be_studied :

Text file, Text file

label_translation :

Label Translation, Label Translation

sulci_label_translation_siGraph :

Text file, Text file

left_imprint_mesh :

MESH mesh, MESH mesh

left_imprint_texture :

Texture, Texture

left_imprint_bottom :

NIFTI-1 image, NIFTI-1 image

left_imprint_hull_junction :

NIFTI-1 image, NIFTI-1 image

right_imprint_mesh :

MESH mesh, MESH mesh

right_imprint_texture :

Texture, Texture

right_imprint_bottom :

NIFTI-1 image, NIFTI-1 image

right_imprint_hull_junction :

NIFTI-1 image, NIFTI-1 image

Talairach_transform :

Transformation matrix, Transformation matrix

subjects_hierarchy :

Text file, Text file