Computer assisted mapping in quantitative analysis of cerebral positron emission tomograms

Area of Science:

• Neuroimaging
• Medical Physics
• Computer Science

Background:

• Quantitative analysis of brain CT scans is crucial for neurological research.
• Existing methods for brain mapping can be time-consuming and lack reproducibility.
• Individual anatomical variations pose challenges for standardized analysis.

Purpose of the Study:

• To develop an interactive FORTRAN program for time-saving, reproducible brain CT mapping.
• To adapt the program for positron emission tomography (PET) using the [18F]-2-fluoro-2-deoxy-D-glucose (FDG) method.
• To ensure flexible adaptation to individual anatomical differences.

Main Methods:

• A two-step mapping procedure: geometric contour definition followed by activity-based delimitation.
• Interactive control allowing user-guided corrections.
• Storage of mapping parameters in a standard file for reproducibility and comparability.

Main Results:

• Demonstrated program scope and functions with seven normal volunteers.
• Achieved high, largely user-independent reproducibility.
• Numerical results align with previous manual mapping studies.
• Good agreement between gray and white matter discrimination and theoretical predictions.

Conclusions:

• The developed program offers an efficient and reproducible method for quantitative brain CT analysis.
• It successfully integrates with FDG-PET imaging.
• The program's adaptability makes it suitable for various CT techniques and individual anatomical needs.