Nearest Neighbor Method to Estimate Internal Target for Real-Time Tumor Tracking

Area of Science:

• Medical Physics
• Radiotherapy
• Image-guided therapy

Background:

• Respiration-induced tumor motion poses challenges in radiotherapy.
• Accurate tumor motion tracking is crucial for effective treatment delivery.

Purpose of the Study:

• To propose and validate a nearest neighbor estimation method for tracking respiration-induced tumor motion.
• To assess the estimation and prediction accuracy of the proposed method.

Main Methods:

• Utilized simultaneously collected motion traces of external surrogates and internal targets.
• Employed nearest neighbor search in external space to determine concurrent targets in internal space.
• Validated the method on 71 cases from open-access databases and compared it with existing methods.

Main Results:

• Achieved a root-mean-square error of less than 3 mm in 69 out of 71 cases.
• Demonstrated superior estimation accuracy compared to polynomial models.
• Showcased good prediction performance for tumor motion.

Conclusions:

• Nearest neighbor estimation is effective for tracking respiration-induced tumor motion.
• The method is modality-independent, requires only one critical parameter, and is easily implementable in clinical settings.
• The method shows potential for broad applications in image-guided radiotherapy.