Audiovisual biofeedback improves motion prediction accuracy.

Area of Science:

• Medical Physics
• Radiotherapy Technology
• Biomedical Engineering

Background:

• Motion prediction accuracy in radiotherapy is crucial for managing system latency.
• Patient respiratory patterns often exhibit irregularities, hindering precise motion prediction.
• Audiovisual (AV) biofeedback has demonstrated potential in reducing respiratory motion irregularities.

Purpose of the Study:

• To evaluate the effectiveness of AV biofeedback in improving the accuracy of respiratory motion prediction.
• To test the hypothesis that AV biofeedback enhances predictive modeling for radiotherapy motion management.

Main Methods:

• Implemented an AV biofeedback system with real-time respiratory data acquisition (abdominal and diaphragmatic).
• Collected respiratory signals from 15 healthy subjects under guided and unguided breathing conditions.
• Utilized a kernel density estimation algorithm for motion prediction and quantified error using Root Mean Square Error (RMSE).

Main Results:

• AV biofeedback improved prediction accuracy in 69% of abdominal wall and 78% of diaphragm motion predictions.
• Average RMSE reduction was 26% for abdominal motion and 29% for diaphragm motion (p < 0.001) with AV biofeedback.
• Statistical analysis confirmed the significant improvement in prediction accuracy.

Conclusions:

• AV biofeedback effectively reduces respiratory irregularities, leading to enhanced motion prediction accuracy.
• Improved prediction accuracy facilitates more efficient motion management in radiotherapy, mitigating system latency challenges.
• This study provides the first evidence of AV biofeedback's benefit for radiotherapy motion prediction accuracy.