Dual-energy (DE) imaging was introduced by R. E. Alvarez and A. Macovski in 1976.(1). DE is an imaging modality which involves acquiring images at two distinct energy levels. The information acquired can be used for enhanced material discrimination. The technique leverages variations between materials in material attenuation at different energy levels, allowing for material discrimination and identification. It can be used to detect specific materials against uneven, cluttered backgrounds.
Potential applications for dual energy-imaging:
security applications
food inspection
breast imaging
chest imaging
dual-energy CBCT for radiotherapy
non-destructive testing (NDT)
The Spectrum Logic/ISDI team worked with UCL on a study which has just been published in Journal of Applied Physics ‘Design and fabrication of a sandwich detector for material discrimination and contrast cancellation in dual-energy based X-ray imaging’
The team developed a ‘Sandwich Detector’ which consists of two detectors one placed on top of the other:
Dual energy data can be acquired by either dual- or single-shot exposure techniques. In dual exposure, kVp switching can be used to acquire low energy (LE) and high energy (HE) sequentially.
In the "Sandwich Detector," the front detector captures low-energy X-ray photons, while the back detector detects high-energy photons. Thus, two automatically registered images are acquired simultaneously, improving imaging speed and avoiding motion artifacts. However, if this approach is not optimized, it can result in poor energy separation and excessive quantum noise.
The aim of the UCL/Spectrum Logic/ISDI study is to determine whether optimization can be implemented based on the current developments in complementary metal oxide semiconductor (CMOS) advanced pixel sensor (APS) technology.
Sandwich Detector construction:
To accomplish this objective, a multi-layer energy integrating detector was built by stacking two CMOS APS sensors coupled with scintillators of appropriate thickness. A copper plate was inserted between the two detectors to act as a filter and to increase spectral separation. High Z materials enable larger spectral separation with a lower filter thickness, and the team chose Cu as it is stable, non-toxic, cost-effective, and easy to source at reasonably high purity.
(iv) Experimental set up
To measure the effectiveness of the “sandwich” detector, material discrimination and contrast cancellation techniques were applied to real data; at relatively low X-ray energies (70 kV - more details in Fig iv) with low Z target materials. This choice was made with prospective applications in breast imaging and food inspection in mind. The detector design was optimized for these specific low-energy spectral characteristics.
All experimental studies conducted involved varying the thickness of the copper filter layer and measuring the effectiveness of the sandwich detector at each thickness (0mm, 0.25mm and 0.5mm). When χ2 tests were performed, the 0.25mm thickness Cu performed the best under the considered spectral energy, as was predicted by the model. This supports the reliability of the model in predicting the optimisation of the thickness of the Cu layer, and so also supports the model's ability to predict the optimal choice for the thickness of the top scintillator, which was unable to be tested experimentally for practical reasons.
Although low energy ranges can be seen as a limitation of the study, the team concluded that the optimization process could be adapted and detectors optimised for different applications in future. The study states that further research at higher energy ranges will be undertaken and cite papers already published on dual energy detection at high kV for both security and cargo inspection (2) (3).
The UCL/ISDI/Spectrum Logic study focuses on the use of CMOS APS sensors, and the team conclude by highlighting other options that could be considered. One option is state-of-the-art single photon counters, which can use thresholding capabilities to perform spectral separation without the need for filters. Another development is detectors based on perovskite materials (4) which is a very active field of research.
Find out more about ISDI and Spectrum Logic’s collaboration with UCL by reading this interview with Rimcy Palakkappilly Alikunju from UCL on the ISDI website.
References:
1. R. E. Alvarez and A. Macovski. “Energy-selective reconstructions in X-ray computerised tomography”
https://iopscience.iop.org/article/10.1088/0031-9155/21/5/002
2. K.Wells and D. A.Bradley , “A review of X-ray explosives detection techniques for checked baggage”.
https://doi.org/10.1016/j.apradiso.2012.01.011
3. D. Velayudhan, T. Hassan, E. Damiani, and N. Werghi, “Recent advances in baggage threat detection: A comprehensive and systematic survey” https://doi.org/10.1145/3549932
4. A. Datta, Z. Zhong & S. Motakef. “A new generation of direct X-ray detectors for medical and synchrotron imaging applications” https://www.nature.com/articles/s41598-020-76647-5