Scan Path Optimization of Discrete Scanning Based on HE-GA in Proton Therapy

Scan path optimization in proton therapy requires efficient intelligent optimization algorithms. In this study, we discuss the optimization of the genetic algorithm (GA) for scan path optimization from three aspects: generating initial paths using heuristic algorithms, optimizing the genetic parameters of GA, and incorporating an elite preservation strategy. We propose a heuristic elite-genetic algorithm (HE-GA) and compare it with commonly used simulated annealing (SA) and GA, finding that HE-GA demonstrates higher accuracy, better stability, and faster computation. Using HE-GA under discrete scanning, we analyze tumor cases under different scanning dead times and observe that optimizing the total scanning time by reducing the scanning dead time can lead to improvements. This enables repetitive scanning with the same duration as the zigzag scanning pattern, providing better protection for patients' critical organs.