Statistical Shape Model of Craniosynostosis Patients with 100 Instances per Pathology

A statistical shape model constructed from 3D surface scans of craniosynostosis patients, created by Matthias Schaufelberger of Karlsruhe Institute of Technology. It includes a full shape model, a texture model, and submodels for four classes of cranial suture fusion and a control group. The dataset provides 100 mesh instances per pathology class for analysis and synthesis.

Use Cases

• Classifying cranial pathologies based on 3D shape features described in the model.
• Synthesizing new 3D head models for surgical planning or simulation using the provided statistical information.
• Integrating shape statistics into existing pediatric head models like the Liverpool-York child head model.
• Training machine learning models for automated diagnosis of scaphocephaly, trigonocephaly, and plagiocephaly.

Strengths

• Includes 100 mesh instances for each of the four pathology classes and the control group.
• Model is compatible with the Liverpool-York child head model, using the same vertex order and IDs.
• Provides multiple model types: a full shape model, a texture model, and specific submodels for different suture fusions.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Last update date is unknown; freshness unverified.

Provenance

Source

Karlsruhe Institute of Technology

Collection Method

Constructed from 3D surface scans of craniosynostosis patients.