SPIE Startup Challenge 2015 Founding Partner - JENOPTIK Get updates from SPIE Newsroom
  • Newsroom Home
  • Astronomy
  • Biomedical Optics & Medical Imaging
  • Defense & Security
  • Electronic Imaging & Signal Processing
  • Illumination & Displays
  • Lasers & Sources
  • Micro/Nano Lithography
  • Nanotechnology
  • Optical Design & Engineering
  • Optoelectronics & Communications
  • Remote Sensing
  • Sensing & Measurement
  • Solar & Alternative Energy
  • Sign up for Newsroom E-Alerts
  • Information for:

SPIE Photonics West 2017 | Register Today

SPIE Defense + Commercial Sensing 2017 | Call for Papers

Get Down (loaded) - SPIE Journals OPEN ACCESS


Print PageEmail PageView PDF

Biomedical Optics & Medical Imaging

A modular high-resolution photon-counting x-ray detector

A modular device that can detect and characterize the energies of individual x-ray photons could prove broadly useful in computed tomography and digital radiography.
29 December 2010, SPIE Newsroom. DOI: 10.1117/2.1201011.003217

For over a century, x-ray imaging detectors (whether film or digital) have formed images by integrating x-ray interactions over a finite acquisition time without regard to event energy or number of events. While the benefits of these ‘energy-integrating’ detectors are well established, their performance—especially in terms of high contrast ratios— is suboptimal. Consequently, the next generation of x-ray detectors for computed tomography (CT) and digital radiography must be capable of counting individual photons and characterizing (and even recording) their energies. In the case of medical imaging, increased sensitivity at low energy will allow reduced radiation dosage to patients and provide contrast ratios equivalent to those obtained by other imaging methods, while improving spatial resolution and reducing noise for the same quantum efficiency. The advanced energy capability will also enable reconstructions with fewer beam-hardening and minimal ghosting artifacts due to lag, and carry the potential for single-exposure, multiple-energy imaging.1

In response to these needs and opportunities, we are developing a family of modular, highly configurable photon-counting, energy-discriminating, high-resolution imaging devices based on a novel combination of cadmium zinc telluride (CdZnTe) semiconductor radiation sensors and high-resolution custom application-specific integrated circuits. Our goal for our Advanced Photon Counting DetectorTM (APCD) module is to combine these two high-performance components into small (e.g., 1×1cm2 and 2×2cm2) modules that tile seamlessly, forming detectors of arbitrary size and 2- or 3D shape (see Figure 1). APCD components and APCD-based detectors are being designed for critical, highly demanding applications such as medical x-ray CT and munitions inspection, as well as for medical and nonmedical x-ray CT and radiography in general. We intend APCD devices to be used to read out other solid-state sensors in addition to CdZnTe. Finally, the APCD design is also being upgraded to read out scintillators.

Figure 1.(a) Cadmium zinc telluride (CdZnTe) or CdTe sensors will be bump-bonded to Advanced Photon Counting Detector (APCD) application-specific integrated circuit (ASIC) CMOS chips to form APCD detector modules (e.g., a 2×2cm2module of 80×80 pixels). (b) Modules may then be tiled to form APCD detectors, continuous pixel arrays of arbitrary size (e.g., 4×4 modules, 8×8cm2, 320×320 pixels). To do so, modules will be mechanically attached individually to printed circuit motherboards supplying power and containing control and readout electronics. Modules may be individually removed for service.

Each APCD module consists of a monolithic sensor (e.g., CdZnTe) of appropriate thickness to absorb x-rays with high efficiency, coupled to our CMOS APCD chip, incorporating at least 40×40 pixels of (typically) 250×250μm2 per pixel or smaller. Each individual pixel will independently support a data rate high enough to effectively measure individual source x-rays, even at high flux. APCD modules will mount mechanically on a motherboard containing the necessary power, ground, control, and readout electronics, and can be individually removed and replaced for service.

For the given applications, high typical x-ray count rates of 106–108 photons/mm2/s, combined with the need for high detection efficiency, require the development of detector structures that can simultaneously provide high spatial resolution, fast response, and efficient x-ray absorption over an energy range of 28kVp (kilovolt peak, typical for mammography) to 140kVp (adequate for whole-body CT). The APCD detector module is designed to deliver spatial resolution of 125–250μm per pixel and energy resolution of better than 3% full width at half-maximum at 120keV, with five software-selectable energy intervals for binning counts during acquisition, and with at least three-side abuttable modularity for organization of APCD detector modules into large detectors. Moreover, all of the necessary electronic circuitry for each APCD pixel fits within the space of that one pixel, and still allows sensing and counting well in excess of 0.5×107 x-ray photons per 250μm2 pixel per second.

Figure 2.CMOS APCD (mounted for testing) and CdZnTe pixel array prototypes, each consisting of 19×19 pixels of 250×250μm2per pixel on a 5.5×5.5mm2footprint.

Each APCD pixel is designed to acquire the charge of each event from its corresponding CdZnTe pixel, classify the total energy of each incoming charge packet into one of five software-selectable energy intervals, and count (histogram) every charge packet by energy. Figure 2 shows a photograph of APCD and CdZnTe prototypes. Both consist of 19×19 pixels of 250×250μm2 per pixel on a 5.5×5.5mm2 footprint. Every APCD pixel includes a 60×60μm2 pad needed for bump bonding to a CdZnTe contact, a preamplifier, a leakage-current subtraction circuit, an auto-zeroed programmable gain stage, five comparators, a variable-delay reset circuit, and five 16-bit counters (see Figure 3). All the required analog circuitry and operation is confined to each individual pixel.

Figure 3.A single-pixel layout. The pixel is 250×250μm2. The left-hand side contains the analog components, including (top to bottom) the programmable feedback capacitor array, the input pad, the input amplifier, the five comparators, and the reset delay circuit. The right-hand side contains the digital components, including the five 16-bit counters and readout circuitry. Analog and digital power are kept separate to improve noise immunity.

Detailed results available elsewhere2,3 from our recent work in characterizing and evaluating the prototype APCD components show the simulation of the clock circle of a complete self-time reset operation of both the analog and digital part of the chip. We describe the time intervals required for the completion of an entire process, including from the point at which the pulse arrives at the preamplifier until all the comparators fire, together with the reset delays set at both preamplifier and comparator, are described. The reset delays to ensure the completion of integration and of charge erasure, as well as the auto-zeroing function, can be independently optimized. In simulation, using the highest-gain setting, the pixel can integrate, count, and reset full-magnitude signals at a rate of 5MHz. For smaller signals or with lower gain settings, a rate of 10MHz is possible.

Because of their superior performance vis-à-vis spatial resolution, signal-to-noise ratio, and dynamic range, and the potential for using single-exposure multiple-energy imaging to improve quantitative accuracy for contrast-enhanced imaging and to allow determination of material density in scans, APCD detectors will be equally well suited for x-ray inspections at airports/borders and for medical imaging. The improved quantitation in preclinical (small animal) imaging, where anatomical features are extremely hard to visualize, should also fuel pharmaceutical and medical research in new drug development and for anticancer therapies. Similarly, the fields of basic biological research using synchrotron radiation, basic physics studies, crystallography, and space research, each of which encompasses many applications, should all benefit from this development.

Our future work will be directed toward improving the performance of the APCD ASIC and developing new designs and different contact patterns of CdZnTe sensors. These will eventually lead to achieving faster device operation and a greater ability to accommodate different pixel and module sizes. The APCD ASIC may also be modified by decreasing or increasing the number of comparators and counters (bins) to accommodate application needs and pixel size constraints.

We thank the US Department of Defense, the National Institutes of Health, and the Department of Energy for supporting this work through grants W15QKN-09-C-0132, 1 R43 CA139683-01, and 91185S09-I (40b), respectively.

Vivek V. Nagarkar, Georgios Prekas, Steven Cool, Radia Sia, Stuart Kleinfelder
Radiation Monitoring Devices Inc.
Watertown, MA