Head and neck lymph node region delineation with 3-D CT image registration

Area of Science:

• Medical Imaging
• Radiation Oncology
• Computational Anatomy

Background:

• Accurate delineation of the target volume is crucial for effective radiation therapy.
• Current manual contouring of tumor spread by radiation oncologists is labor-intensive and challenging.
• Estimating microscopic disease spread remains a significant hurdle in treatment planning.

Purpose of the Study:

• To develop automated methods for defining radiation therapy target volumes.
• To reduce the workload of radiation oncologists in treatment planning.
• To improve the accuracy of delineating microscopic tumor spread.

Main Methods:

• Developing methods for automatic selection and adaptation of standardized tumor spread regions.
• Utilizing lymph node locations from a reference case.
• Employing image registration techniques, including deformable transformations with "mutual information" optimization.
• Integrating anatomic knowledge-based methods.

Main Results:

• Image registration techniques show promise for target volume delineation.
• Anatomic knowledge-based methods are necessary to supplement image registration.
• The developed methods aim for a clinically acceptable match for tumor spread.

Conclusions:

• Automated target volume delineation holds potential to enhance radiation therapy success.
• Combining image registration with anatomic knowledge is key for clinical applicability.
• Further development is needed to achieve clinically acceptable automated contouring of microscopic disease spread.