[Simulation of lung motions using an artificial neural network].

Area of Science:

• Medical Physics
• Radiotherapy
• Computational Biology

Background:

• Accurate lung radiotherapy requires understanding patient lung motion.
• Lung motion, driven by breathing, causes displacements in the clinical target volume (CTV).
• Precise tracking of CTV displacements is crucial for effective radiotherapy.

Purpose of the Study:

• To present a feasibility study of a novel method for simulating lung motion at all breathing phases.
• To improve the accuracy of lung radiotherapy by understanding lung motion.
• To enable the tracking of clinical target volume (CTV) displacements induced by breathing.

Main Methods:

• An artificial neural network (ANN) was employed to learn lung motion patterns.
• The ANN was trained on data from three patients, utilizing over 600 plotted points in a specific lung area.
• The model simulates lung motion for new patients using only initial and final breathing data.

Main Results:

• The developed method achieved a promising average accuracy of 1mm.
• The simulation was performed with a spatial resolution of 1 × 1 × 2.5mm³.
• Initial results indicate the feasibility and potential of the ANN approach.

Conclusions:

• It is feasible to accurately simulate lung motion using an artificial neural network.
• The study demonstrates the potential for improving lung radiotherapy accuracy.
• Future work includes enhancing accuracy by incorporating more patient data and covering the entire lungs.