Biomedical Engineering, Vol. 46, No. 5, January, 2013, pp. 194198. Translated from Meditsinskaya Tekhnika, Vol. 46, No. 5, Sep.Oct., 2012, pp. 2126. Original article submitted April 16, 2012.

A Matrix GalliumArsenide Detector for Roentgenography A. P. Vorobiev1*, S. N. Golovnya1, S. A. Gorokhov1, V. V. Parakhin1, M. K. Polkovnikov1, G. I. Ayzenshtat2, M. A. Lelekov2, O. B. Koretskaya3, V. A. Novikov3, O. P. Tolbanov3, A. V. Tyazhev3, D. V. Borodin4, and Y. V. Osipov4

A matrix galliumarsenide detector for roentgenography and nondestructive monitoring is described in this work. The detector contains 128 × 128 sensitive elements with 50μm pitch, an analog multiplexor, and a system for information acquisition. The detector is intended for use in low or mediumdose medical Xray apparatuses. It can also be used for nondestructive defectoscopy.

Interest in digital information systems, which are superior to conventional analog systems in efficiency of information acquisition, storage, and transmission, is gradually increasing. This requires further development of detection technology. Pixel microassembly multichan nel semiconductor systems with matrix detectors are promising detection systems for information acquisition and processing. Such microassemblies have sufficiently large areas and allow for realtime imaging. The goal of this work was to describe a matrix galliumarsenide detector for roentgenography composed of 128 × 128 sensitive elements with 50μm pitch and an analog mul tiplexor. The working characteristics of the detector are also given [1]. The microassembly is a prototype of an X ray image processing system that is presently being devel oped at the Institute of High Energy Physics, Protvino [2].

thickness of this area is 430450 μm, which provides high efficiency of Xray detection. The analog multiplexor is a highly integrated circuit that provides effective parallel integration of currents of the sensitive elements and serial transmission of signals. A specific feature of the circuit is that it implements the function of distraction of the constant component before integration. The subtracted value can be corrected according to the temperature and applied voltage using a 5bit divider. This correction makes it possible to avoid overload of the integration capacitor at high currents of the detector. The circuit also has a window function, which is used to determine the multiplexor size and posi tion. The minimal window size (4 × 4 pixels) can be

TABLE 1. Multiplexor Specifications 128 × 128

Multiplexor format

Electronics of the Detector Microassembly

Cell size, μm Charging capacity, е

The microassembly circuit contains a semiconductor detector based on highresistivity gallium arsenide (GaAs) [35]. The 128 × 128 sensitive elements with 50 μm pitch cover the detector area of 6.4 × 6.4 mm2. The

50 × 50 −

Number of outputs Integration time Signal frequency, MHz

More than 2⋅107 1 (differential) Equal to frame detection time Typical (7)

Maximal frametoframe frequency, kHz: at maximum frame size (128 × 128) at minimum frame size (4 × 4)

1

Institute of High Energy Physics, Protvino, Moscow Region, Russia; Email: [email protected] 2 Federal Research Institute of Semiconductor Devices, Tomsk, Russia. 3 Siberian Physical Technical Institute, Tomsk, Russia. 4 RTK IMPEX, Moscow, Russia. * To whom correspondence should be addressed.

Output signal range, V Scattered power at frame size 128 × 128, mW

194

0.5 250 More than 2 35

00063398/13/46050194 © 2013 Springer Science+Business Media New York

GalliumArsenide Detector for Roentgenography

increased by a factor of 4. The window position on the surface of the matrix can be determined accurate to 4 pix els. The multiplexor specifications are given in Table 1. Application of time signals to the analog input of the multiplexor generates voltage levels proportional to the integration capacitor charge. The multiplexor generates new frame and new string signals that are used for image reconstruction. The detector and multiplexor are placed in a ther mostatic unit. A Peltier element with working power of 53 W and temperature gradient of 72°С is used as the thermostat. The cold surface of the Peltier element is made of aluminum. The circuit plate contains a hole for air passage. The hightemperature element of the Peltier circuit is cooled using a DC fan. The outer walls of the unit are insulated using foam plastic. The internal temperature of the unit is controlled using a TMR36 thermal sensor. The error of the temper ature control is 0.1°C.

Detector bias voltage source

XRay tube power source

195

Data Acquisition System Analog multiplexor data are collected using a special data collection system. The system consists of a coupling amplifier, a pipeline ADC, a static memory for the results of measurement, a control unit, and an optical interface. The FT245BM microcontroller (FTDI, Great Britain) [6] is based on an 8bit parallel port and a standard serial USB port. The system identification unit is based on an EEPROM microcircuit. The FT245BM microcontroller is powered from the USB port. The system also contains two bipolar power sources for analog and digital circuits and a stabilized power source with a microammeter for monitoring the dark current of the detector. The schemat ic diagram of the data acquisition unit is shown in Fig. 1. A RAPAND70 Xray tube (VNIIA) was used [7]. The Xray tube could be controlled with a remotecontrol panel or via PC. The data acquisition system was placed into a leadcoated box providing protection from Xray radiation.

XRay tube control panel

Microammeter for monitoring dark current of detector

Cooling system power source

Amplifier 128 × 128 matrix ADC 12 bit 10 MHz

Cooling system and thermostabilization

Analog circuit power source

Sampling and storage control unit

Static memory 128 kB × 12

Digital circuit power source

Agilent 33120A digital generator

OS

Fig. 1. Schematic diagram of data acquisition unit.

196

Vorobiev et al.

Software The MATRIX software was developed to provide data reading and processing. The software is for LabWindows (National Instruments, USA) [8]. A general view of the main window of the software is shown in Fig. 2. The control elements of the MATRIX software con tain 5 subwindows used for: graphical data presentation; image correction and smoothing; presentation of USB bus parameters; control of Xray anode voltage and current; working temperature control. The MATRIX software pro

vides connection to a USB storage device and reading, presentation, and harddisk storage of data retrieved from the device. The interaction with the FT245BM micro controller is performed using the FTD2XX interface. Lowcontrast images can be corrected using several calibration methods to improve the contrast. Correction is based on the use of two images: an image obtained without Xray exposure of the detector and a background image obtained by uniform exposure of the detector to the radia tion of the Xray tube. Two correction coefficients are cal culated: R (shift with respect to zero) and F (scaling factor).

Fig. 2. General view of the main window of the software for data reading and processing.

Fig. 3. Image of test pattern with 5.5 pairs of lines per mm.

Fig. 4. Image of test pattern with 8 pairs of lines per mm.

GalliumArsenide Detector for Roentgenography

197

Fig. 5. Image of test pattern with 9 pairs of lines per mm.

Fig. 6. Image of test pattern with 10 pairs of lines per mm.

Fig. 7. Test object with contrast sensitivity of 1.5%.

Fig. 8. Test object with contrast sensitivity of 1%.

As the image is displayed on the PC monitor, it can be subjected to a number of special manipulations such as variation of image contrast, statistical processing, and false coloring. These additional functions increase the efficiency of the MATRIX software.

pattern with 10, 9, 8, and 5.5 pairs of lines per mm. The results of the measurement are given in Figs. 36. Unfortunately, the small size of the detector area makes it impossible to simultaneously present the image of the test pattern and its number. As noted above, the detector pitch is 50 μm. The limiting value of the spatial resolution is 10 pairs of lines per mm. The results of this work demonstrated that this limit was not attained (Fig. 6), which can be due to charge separation between several pixels of the detec tor.

Results Measurement of spatial resolution. The spatial reso lution of the detector was measured using a standard test

198

Measurement of contrast sensitivity. The contrast sen sitivity was measured using a test object and a special filter. The test object/filter thickness ratio gives the contrast value. During the measurement, the test object was placed above the detector and covered with the filter. The standard test set included a set of plates made of highpurity alu minum and used as 3, 2, 1.5, 1, and 0.5% filters. The test object comprised aluminum samples of circular, triangular, and square shape placed between two thin sheets of plastic. A polished bar of very highpurity aluminum (>99.99%) was used as the main filter. The anode voltage was 70 keV; anode current, 7 mA. According to testing recommenda tions (Standard 012204 [9]), the object–focus distance was 0.5 m. The results obtained in objects with contrast sensitivities of 1.5 and 1% are shown in Figs. 7 and 8. The results demonstrate that the contrast sensitivity of the matrix detector described above is at least 1%.

Conclusion The results of this work are only a first step in intro ducing digital filmless radiographic systems based on matrix semiconductor GaAs detectors into the domestic

Vorobiev et al.

medical practice. The problems of interconnection of matrix detectors, selection of optimal working parame ters, and scanning and data processing are yet to be solved. The results of this work demonstrate that certain techniques based on the use of galliumarsenide detectors hold considerable promise for practical application. REFERENCES 1. D. V. Borodin and Yu. V. Osipov, Prikl. Fiz., No. 6, 9899 (2003). 2. www.ihep.su 3. A. P. Vorobiev, S. N. Golovnya, S. A. Gorokhov, et al., Nov. Prom. Tekhn., No. 2, 49 (2005). 4. A. V. Tyazhev, D. L. Budnitsky, O. B. Koretskaya, A. I. Potapov, O. P. Tolbanov, L. S. Okaevich, V. A. Novikov, and A. P. Vorobiev, Nucl. Instrum. Meth. Phys. Res., 509, 3439 (2003). 5. G. I. Aizenshtat, E. N. Ardashev, A. P. Vorobiev, A. I. Ivashchenko, O. B. Koretskaya, O. P. Tolbanov, A. V. Tyazhev, and A. V. Khan, Elektron. Prom., No. 2/3, 3236 (2002). 6. www.ftdichip.com 7. www.vniia.ru 8. www.ni.com 9. Standard 012204 VNIIIMT “Image Receivers Used in Diagnostic XRay Apparatuses with Digital Image Detection: Parameters and Characteristics of Image Quality and Methods and Devices for Its Determination” [in Russian], VNIIIMT (2004).