Preliminary study of a clinical electron beam using highly pixelated CMOS Image Sensors
Introduction
The precise and accurate determination of the flux of intense ionizing radiation beams is of paramount importance especially when they are used to treat tumours. Many efforts are put into measuring with 1% uncertainty the energy deposition both in the tumour and the healthy tissues during a radiotherapy session. Many different types of detectors have been used up to now for such measurements. The increasing complexity of Treatment Plans and the reduction of the ionizing beam size makes very difficult the measurement task.
Thanks to the developments of the Complementary Metal–Oxide Semiconductor (CMOS) technology, the single pixels of CIS devices are reducing in size and increasing in complexity of the on-pixel and peripheral electronic. CIS devices are optimized to convert visible light into a digital image to record video or photo, hence this implies a very small pixel size, to increase the picture definition remaining within the standard form factors of the recording devices (usually below 1 cm2). Recently some investigation on the capability of CIS to detect ionizing radiation rather than visible light has demonstrated their efficiency in this task, especially for charged particles like electrons (Turchetta et al., 2001, Battaglia et al., 2008, Faruqi et al., 2005, Evans et al., 2005, Battaglia et al., 2009, Servoli et al., 2008, Servoli et al., 2010, Servoli et al., 2011, Castoldi et al., 2015, Perez et al., 2016). This is due to the extremely small Equivalent Charge Noise (ECN) that can be as low as few e for the standard 33 ms integration time, while the signal released by a minimum ionizing particle crossing the sensor is in the range of hundreds e, even for the smallest thickness of the epitaxial region, namely 2- (Servoli et al., 2013). Hence a Signal/Noise exceeding 30 could be easily obtained and flux measurement based on the identification of each electron on the matrix could be performed. The limitation of this method lies in the possible superposition of signals when the flux increases. There are two types of effects in this case: a pixel may have a superposition of signals from different electrons, resulting in a signal over the threshold (fake electron); two electrons may hit two adjacent pixels resulting in just one cluster with one relative maximum (reduced detection). Overall effect is a sublinear behaviour that requires a non-linear calibration. A possible method attaining a precision better than 1% in the flux determination, even in the presence of intense fluxes of charged particles at therapeutic accelerators, has been proposed (Servoli and Tucceri, 0000, Servoli and Tucceri, 2016). In this work we will explore the capability of such highly segmented devices to characterize precisely ionizing radiation beams, extracting the information on their variability in time and homogeneity in space.
Section snippets
Experimental setup and reference flux measurement
Aptina Imaging (http://www.aptina.com/) is one of the leading world suppliers of CIS, its devices are bonded to a standard sensor headboard read out by a custom DAQ system (Demo2 board) based on a Xilinx FPGA powered and controlled by a PC via a USB connection. In this work we have exposed a sensor MT9T031 to 10 MeV electron beam delivered by an Elekta e-LINAC in use at Santa Maria Hospital (Terni, Italy), mounting a 3 × 3 cm2 collimator. Specific features of the sensor are pixel size,
Determination of uncertainty factors
The uncertainty with respect to the reference flux value is due to multiple factors, among which the stability of the beam in time, the spatial uniformity over the sensor surface and the intrinsic sensor uncertainty in the measurement:
To disentangle the different contributions we have developed a method based on both the very high spatial segmentation of the sensor and the multiple measurements of the flux in time.
In the following of
Conclusions
Currently there are not sensors capable of measuring the beam flows with such small pixelization and with such precision due to the difficulties of counting the single radiation-detector interactions. CIS, thanks to their high spatial segmentation, after a suitable non-linear calibration procedure, could measure high fluxes of charged ionizing radiation up to tens of MHz/cm2. We have developed a method which uses one irradiation session lasting few minutes to extract beam characteristics like
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
References (15)
- et al.
CMOS pixel sensor response to low energy electrons in transmission electron microscopy
Nucl. Instrum. Methods A
(2009) Tracking and vertexing with a thin CMOS pixel beam telescope
Nucl. Instrum. Methods A
(2008)CMOS active pixel sensors for ionising radiation
Nucl. Instrum. Methods A
(2005)- et al.
Direct single electron detection with a CMOS detector for electron microscopy
Nucl. Instrum. Methods A
(2005) Particle detection and classification using commercial off the shelf CMOS image sensors
Nucl. Instrum. Methods A
(2016)- et al.
Use of standard CMOS imagers as position detectors for charged particles
Nuclear Phys. B Proc. Suppl.
(2011) A monolithic active pixel sensor for charged particle tracking and imaging using standard VLSI CMOS technology
Nucl. Instrum. Methods A
(2001)