Radiation imaging X-ray detector imaging colorful images

The human eye has blue, green, and red cone cells. Through the synergistic response of tricolor cone cells to visible light in the range of 400-780 nm, we see a colorful world. However, in the shorter wavelength X-ray band (1 pm~10 nm), X-ray detectors used as "eyes" generally use charge integral detection, which does not have wavelength and energy resolution capabilities. Therefore, common X-ray films, CT examinations, etc. in daily life can only obtain black and white grayscale imaging.
In fact, the photon energy of X-rays carries a lot of information. The attenuation of rays depends on the energy of emitted photons and the density and composition of the object being penetrated. For example, low-density organic substances such as human soft tissues almost fully transmit high-energy hard X-rays, and only have good imaging contrast under low-energy rays less than 30 keV; High density bones, on the other hand, fully absorb low-energy rays and require high-energy rays ranging from 80 keV to 140 keV to penetrate imaging.
In fact, the photon energy of X-rays carries a lot of information. The attenuation of rays depends on the energy of emitted photons and the density and composition of the object being penetrated. For example, low-density organic substances such as human soft tissues almost fully transmit high-energy hard X-rays, and only have good imaging contrast under low-energy rays less than 30 keV; High density bones, on the other hand, fully absorb low-energy rays and require high-energy rays ranging from 80 keV to 140 keV to penetrate imaging.
In order to obtain more information from high-energy and low-energy X-rays, the development of dual or multi energy resolution X-ray detection is an important research direction in related fields. This type of detector is beneficial for distinguishing substances of different densities, enhancing imaging contrast between organic and inorganic objects, and identifying soft substances (such as rubber, plastic, and soft tissue) with poor radiation absorption. In addition, the detector can also use digital subtraction to extract images of different substances at the same location, such as subtracting bones from chest X-rays to display only the lungs.
The current methods for achieving dual energy X-ray detection include changing the energy of the radiation source and increasing the number of detector layers. The former exposes the object twice in a row to an X-ray source with a changing energy range. In order to reduce the movement between the two exposures (such as heart beating, respiratory movement, etc.), it is necessary to quickly switch the voltage of the X-ray source, which puts very high technical requirements on the equipment. In addition, the radiation dose is inevitably increased during multiple X-ray exposures. The latter becomes a better choice due to its ability to detect under a single irradiation, avoiding the disadvantage of two irradiation cycles. The top layer of the double-layer detector prioritizes detecting low-energy photons, while the bottom layer detects filtered high-energy photons. However, only two layers of energy resolution cannot meet complex imaging requirements, and the commonly used indirect X-ray detection (X-ray visible light electrical signal) limits detection sensitivity, requiring higher radiation doses for clear imaging.
For this purpose, the Wuhan National Optoelectronic Research Center of Huazhong University of Science and Technology, led by Professor Tang Jiang, and the team of optoelectronic devices and 3D integration, Professor Niu Guangda, and others used highly sensitive metal halide perovskite for direct X-ray detection (X-ray → electrical signal), designed a vertical array electrode structure arranged along the direction of X-ray incidence, and achieved a high-energy resolution multi energy X-ray detector under single exposure.
This achievement was published under the title "Vertical matrix perovskite X-ray detector for effective multi energy discrimination" in Light: Science&Applications.
In this article, the researchers used the variation of X-ray photon attenuation depth with photon energy in a vertical array to construct a conversion relationship between X-ray energy and array electrode response in five ranges (Figure 1), and proved that this parameter only depends on the structure and performance of the detector itself, and has universality for different X-ray spectra. Machine learning training on spectral matrix and response matrix datasets can yield the optimal response transformation matrix.

    The researchers used MAPbBr æ perovskite single crystal to prepare a multi energy detector for this structure, and verified the energy resolution of X-rays in the energy range of 30 keV to 70 keV. The perovskite detector in this work has good response linearity and a linear fitting coefficient R ²>  0.99. Linear response is a prerequisite for detectors to accurately distinguish unknown X-ray spectra, ensuring that signal distribution is not affected by radiation intensity. The experiment proved that the distribution of radiation response fluctuates reasonably. As the driving voltage of the radiation source increases from 30 kV to 70 kV, the number of high-energy photons in the radiation increases, and the response ratio of the rear electrodes 3, 4, and 5 increases.

 The response transformation matrix was obtained based on the test dataset and simulated spectral matrix set shown in Figure 2, and the accuracy of this matrix was verified. The detector accurately identified the cutoff energy edge of the X-ray spectrum emitted at operating voltages of 45 kVp, 55 kVp, and 65 kVp, with errors as low as 2.77% (55 kVp) and 2.97% (65 kVp) compared to the simulated spectrum, verifying the effectiveness of the conversion matrix (as shown in Figure 3a).

The principle of using a multi resolution X-ray detector to achieve material differentiation imaging is shown in Figure 3b. Light density substances have higher contrast in front-end electrode imaging, while heavy density substances are clearer in back-end electrode imaging. Digital subtraction of the two imaging groups can achieve material separation. In the article, artificial samples with similar density and elements were used to simulate human bones, muscles, and fat. The three substances in the artificial samples were clearly distinguished and imaged using a multi energy detector in this study (Figure 3c-f).
In summary, this work theoretically and experimentally validates a vertical array perovskite X-ray detector design that can achieve multi energy resolution. It is expected to be used for a new generation of X-ray detection with energy spectrum discrimination, substance differentiation, and enhanced imaging contrast, to "draw colors" to black and white X-ray photos and provide more hidden information in applications such as disease diagnosis.