Retrieving Proton Beam Information Using Stitching-Based Detector Technique and Intelligent Reconstruction Algorithms.

In view of the great need for quality assurance in radiotherapy, this paper proposes a stitching-based detector (SBD) technique and a set of intelligent algorithms that can reconstruct the information of projected particle beams. The reconstructed information includes the intensity, sigma value, and location of the maximum intensity of the beam under test. To verify the effectiveness of the proposed technique and algorithms, this research study adopts the pencil beam scanning (PBS) form of proton beam therapy (PBT) as an example. Through the SBD technique, it is possible to utilize 128 × 128 ionization chambers, which constitute an ionization plate of 25.6 cm, with an acceptable number of 4096 analog-to-digital converters (ADCs) and a resolution of 0.25 mm. Through simulation, the proposed SBD technique and intelligent algorithms are proven to exhibit satisfactory and practical performance. By using two kinds of maximum intensity definitions, sigma values ranging from 10 to 120, and two definitions in an erroneous case, the maximum error rate is found to be 3.95%, which is satisfactorily low. Through analysis, this research study discovers that most errors occur near the symmetrical and peripheral boundaries. Furthermore, lower sigma values tend to aggravate the error rate because the beam becomes more like an ideal particle, which leads to greater imprecision caused by symmetrical sensor structures as its sigma is reduced. However, because proton beams are normally not projected onto the border region of the sensed area, the error rate in practice can be expected to be even lower. Although this research study adopts PBS PBT as an example, the proposed SBD technique and intelligent algorithms are applicable to any type of particle beam reconstruction in the field of radiotherapy, as long as the particles under analysis follow a Gaussian distribution.