Abstract
In clinic, the detection of breast metastasis utilizes the aggregating effect of 99mTc labeled pharmaceuticals in sentinel lymph node (SLN). The doctor hold the gamma counting probe to locate the most radioactive (“hottest”) node (i.e. SLN) during operation and perform the resection. But the gamma counting probe technique is susceptible to noise and lack of depth information. In recent years, many groups devoted to develop an imaging probe suitable for the use in the operating room, hoping to overcome the drawbacks of the gamma counting probe. However, traditional single photon emission computed tomography (SPECT) needs to scan 180 degrees to enable the reconstruction of the three dimensional image. As a result, it is impossible to perform real-time imaging for operation purpose. Our laboratory proposes the DuPECT (dual photons emission computed tomography) system employing the radioisotopes that emit two photons simultaneously. The DuPECT system uses the collimator pairs to restrict the direction of incoming photons and the coincidence circuit to determine whether the detected two photons are emitted in the same decay. The source position is simply the intersecting of the trajectory of the detected photons. The DuPECT system provides three-dimensional spatial information without requiring the rotation of the system. Since the biggest limitation of the DuPECT is its low sensitivity, we try to take advantage of the converging collimators (including the fan-beam collimator and cone-beam collimator) to improve the sensitivity. We utilized the GATE Monte-Carlo simulation (MCS) to validate the feasibility of the DuPECT. However, GATE is time-consuming and can’t simulate the geometry with small focal length. So it is essential to develop a more efficient and flexible geometry-based simulation system. In GATE, the interaction between the photons and collimator takes a large portion of the simulation time. Therefore in this study, we design a new Virtual collimator combined with the SimSET to speed up the simulation. The projection data of the Virtual collimator are compared with GATE to validate the accuracy of the proposed system. Both 99mTc and 75Se are used in the slat collimator and the fan-beam collimator simulation. After demonstrating the feasibility of the Virtual collimator, it was used to construct the DuPECT system. In the study, the optimal focal length of the converging collimator was determined first, and then testing the sensitivity and resolution of fan-beam collimator and cone-beam collimator separately to pair with the slat collimator. According to both slat and fan-beam collimator experimental results, we demonstrate that Virtual collimators are in good agreement with GATE simulation in terms of resolution and sensitivity. In Comparison to GATE, the Virtual collimator improve 2~3 and 20~60 times simulating efficiency for slat collimator and fan-beam collimator, respectively. We conclude that the Virtual collimator boost simulating efficiency and produce the accurate results even on complex geometry. The preliminary results suggest the use of Cone-beam collimator for the DuPECT system as it show better resolution and sensitivity than fan-beam collimator. The system sensitivity is 5.42 cps/kBq with resolution of 2.12*2.12*4.38 mm at x, y, and z direction when the source is placed at the focal point. The results demonstrate that the DuPECT system is feasible for intraoperative detection of the SLN. Although the system is strongly position dependent, we believe that the DuPECT-SLN Probe System will come into clinical trial stage in the future.