Repeated CT scan in improving the reproducibility of grass tumor volume for moving target / 中华放射肿瘤学杂志

ABSTRACT

Objective To find a method to improve the range accuracy of moving target such as peripheral lung tumors, since a single CT snapshot may not be accurate during the treatment process.Methods A simple harmonic motion phantom, embedded with a cube and a circular ball, was used to simulate the tumor motion. Individualized moving targets were scanned 24 times with different amplitudes and frequencies. Then the images were fused from every 1, 2 or 3 sets of CT scans. The GTV volume variation of circular target and the length variation of the cube target along the z axis were contoured and analyzed. Results As motion amplitude increased, the maximum of both circular target volume and cube target length was increased, while the minimum of the factors was decreased. Motion frequency affected the target volume less than amplitude. For a cube target with the length of 3.3 cm at stationary phase, when motion frequencies was 20 and motion amplitude was 2 cm, the maximal length was 2. 4 times of the minimal length (5. 1 cm vs. 2. 1 cm). When it came to the cube target groups fused from every 1,2 and 3 sets of CT scans, the average length and standard deviation were (3.77 ± 1.20) cm, (4.18 ±0. 91)cm and (4.52 ±0. 59) cm, respectively. With the increase of fused scan number, targets became bigger, the standard deviation decreased, and the change of center positions was decreased. Conclusions The motion amplitude, frequency and the number of CT scans are the main factors affecting target definition, though, the optimized scanning phase is not certained. When 4DCT and respiration gating technique are not available,the efficient and practical method to solve this problem is to scan the target three or more times and fuse them in planning system, which will generate a larger, more reproducible GTV volume for moving targets.