Business Context
A radiology department wants to automatically extract 3D surfaces of potential tumors from patient CT scans. The goal is to isolate high-density regions (tumors) from surrounding soft tissue to assist surgeons in pre-operative planning.
About the Set : bioMedImage
Processing and analysis of biomedical images (e.g., DICOM).
Discover all actions of bioMedImage Data Preparation
Simulation of a pre-loaded CT scan table containing a 3D image blob.
Copied!
| 1 | |
| 2 | DATA casuser.ct_scans; |
| 3 | LENGTH _id_ $32 _path_ $255 _image_ binary; |
| 4 | _id_='patient_001'; |
| 5 | _path_='/mnt/ |
| 6 | data/scan01.dcm'; |
| 7 | _image_='000102030405060708090A0B0C0D0E0F'x; |
| 8 | /* Simulated binary blob representing 3D volume */ |
| 9 | RUN; |
| 10 |
Étapes de réalisation
1
Verify input table availability.
Copied!
| 1 | |
| 2 | PROC CAS; |
| 3 | TABLE.tableInfo / TABLE={name='ct_scans', caslib='casuser'}; |
| 4 | |
| 5 | RUN; |
| 6 |
2
Generate 3D surface for tumor density (approx 1200 Hounsfield units) using specific intensity.
Copied!
| 1 | |
| 2 | PROC CAS; |
| 3 | bioMedImage.buildSurface / images={TABLE={name='ct_scans', caslib='casuser'}} outputFaces={name='tumor_faces', replace=true} outputVertices={name='tumor_vertices', replace=true} intensities={1200}; |
| 4 | |
| 5 | RUN; |
| 6 |
Expected Result
The action should successfully identify the isosurface at intensity 1200. Two output tables 'tumor_faces' and 'tumor_vertices' are created, containing the geometric primitives required to render the 3D shape of the tumor.