Purpose: our aim is to develop a portable device for real time personnel dosimetry in IR to perform: i) on line monitoring of staff operations producing an alarm when the dose exceeds a warning level; ii) off line storage of dose measurements in order to correlate them with the specific activities of the staff [1]. The proposed architecture includes a commercial Active Pixel Sensors and each frame carries 600 kB of information. Therefore a data reduction strategy is mandatory to operate the device in real time and at low power. Methods and materials: We have investigated the sensor response to the X-ray radiation scattered by a phantom [2]. First of all we have performed an off-line analysis of collected data at several working settings of the angiographic system, defining two system observables (number of photons detected and the sum of the reconstructed photon signals per second), and studying their capability to serve as dosimetric quantities. Each observable has been extracted using a custom clustering algorithm (algorithm A). TLDs have been used for evaluating the dose at the sensor position and results show a linear correlation among the variables. Then, a simplified and custom algorithm (algorithm B) has been implemented with the goal of retrieving a different dosimetric quantity (E) from a subset of the collected data and being compatible with the real time requirements. Results: To evaluate the performance of the algorithm B we calculated the relative error and compared the behavior of the observable E with the two observables obtained with the data analysis of the algorithm A. The E dosimetric observable of algorithm B and both the number of photons and the sum of reconstructed photon signals are linearly correlated. The relative error of E is around 4% confirming the effectiveness of the proposed approach. Conclusion: The sensor performance as an X-ray radiation detector has been evaluated with a dedicated experimental set-up and dosimetric observables have been assessed from the frames acquired by the sensor using a purposely designed algorithm.
Dosimetric Observables to be used in Active Pixel Sensor based devices for Interventional Radiology Applications
BISSI, LUCIA;PLACIDI, Pisana;CONTI, ELIA;
2013
Abstract
Purpose: our aim is to develop a portable device for real time personnel dosimetry in IR to perform: i) on line monitoring of staff operations producing an alarm when the dose exceeds a warning level; ii) off line storage of dose measurements in order to correlate them with the specific activities of the staff [1]. The proposed architecture includes a commercial Active Pixel Sensors and each frame carries 600 kB of information. Therefore a data reduction strategy is mandatory to operate the device in real time and at low power. Methods and materials: We have investigated the sensor response to the X-ray radiation scattered by a phantom [2]. First of all we have performed an off-line analysis of collected data at several working settings of the angiographic system, defining two system observables (number of photons detected and the sum of the reconstructed photon signals per second), and studying their capability to serve as dosimetric quantities. Each observable has been extracted using a custom clustering algorithm (algorithm A). TLDs have been used for evaluating the dose at the sensor position and results show a linear correlation among the variables. Then, a simplified and custom algorithm (algorithm B) has been implemented with the goal of retrieving a different dosimetric quantity (E) from a subset of the collected data and being compatible with the real time requirements. Results: To evaluate the performance of the algorithm B we calculated the relative error and compared the behavior of the observable E with the two observables obtained with the data analysis of the algorithm A. The E dosimetric observable of algorithm B and both the number of photons and the sum of reconstructed photon signals are linearly correlated. The relative error of E is around 4% confirming the effectiveness of the proposed approach. Conclusion: The sensor performance as an X-ray radiation detector has been evaluated with a dedicated experimental set-up and dosimetric observables have been assessed from the frames acquired by the sensor using a purposely designed algorithm.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
