1. Introduction

Conventional hard X-ray imaging is based on attenuation as contrast mechanism and has been widely applied in medical diagnostics and non-destructive inspections. For soft tissues and other materials made of low-Z elements (such as H and C), however, it is impossible to obtain high-contrast absorption images. The phase factors for these materials are usually three orders of magnitude larger than their absorption factors [1]. Therefore, phase contrast imaging significantly improves the image contrast of these types of objects.

Several types of X-ray phase contrast imaging have been developed recently. They can be categorized into interferometric methods [2,3], free-space propagation techniques [4–7], diffraction enhanced imaging [8–10] and grating interferometry [11–18]. The method using a grating interferometer has many advantages over the others and was first used in applications with synchrotron radiation sources [11–15]. Introducing an absorption grating near the conventional low-brilliance X-ray tube [16–18] enables the grating-based differential phase contrast imaging method, which has potential applications in ordinary laboratories.

The aforementioned grating-based methods use absorption gratings as key optical elements. However, higher X-ray photon energies are required for most clinical and industrial applications, thus, it is essential to manufacture large-area absorption gratings with higher aspect ratios to provide sufficient image contrast for this energy range. The fabricating process for this kind of absorption grating is currently a great challenge, and even state-of-the-art absorption materials used for absorption gratings rely on Au [19,20]. In addition, the transmission properties of absorption gratings lead to deterioration in the image quality. To address these issues, we have proposed a non-absorption grating approach for X-ray phase contrast imaging [21]. To this end, A laboratory-based multi-line X-ray source has been developed to replace the source grating [22–24], which essentially overcomes the limitation of un-ideal transmission from the source grating especially for the case using higher X-ray photon energies (Au can’t absorb X-rays completely), and a structured scintillator has been proposed and developed to replace the combination of absorption analyzer grating and X-ray scintillator and overcome its main drawback: the reduction of the fringe visibility for higher X-ray photon energies [25,26]. In this letter, we give the details of our non-absorption grating approach and demonstrate its primary results for X-ray phase contrast imaging.

2. Experimental setup

The arrangement is shown in Fig. 1 . It consists of a multi-line X-ray source with a spatial period of $p_0$, a phase grating $G_1$ of period $p_1$ with $\pi$ phase shift and an analyzer structured scintillator $G_2$ of period $p_2$. A ladder-shaped structure is fabricated on a tungsten fixed anode by Micro-fabrication techniques. The oil cooling method is used to solve the heat problem. Each step’s length at axial direction is $100\mu m$, and the altitude difference between the adjacent two steps is $31.5\mu m$. The angle between the optical axis and each step’s top surface of the ladder-shaped anode is $6^{\circ }$, thus the apparent pitch of the multi-line source is $42\mu m$, and each apparent X-ray emission line width is $10.5\mu m$. Our ladder-shaped structure can reduce the length of multi-line source along the optical axis, which has a maximized emission area and minimal axial length. Corresponding to the functions of both the conventional X-ray source and source grating, the ladder-shaped multi-line X-ray source generates individually spatial coherent but mutually incoherent line source. Its advantages are obvious: it integrates the source grating into the conventional X-ray tube, improves the mechanical stabilization, increases the fringe contrast especially for higher X-ray photon energies, and provides a cheap and commercially available source with uniform radiation for the entire field of view. For the phase contrast image formation process, the spatial coherence length $l_c$ should satisfy the condition [16]

 

Fig. 1 (Color online) The principles of the non-absorption grating X-ray phase-contrast imaging. The system consists of a ladder-shaped multi-line X-ray source, a phase grating and a structured scintillator.

Download Full Size | PDF

The phase grating has a phase shift of $\pi$. It acts as a beam splitter, diffracts the incoming X-rays into the ± 1st orders and attenuates the zero and higher order components. Over 80% of the X-ray energy is carried by the ± 1st order [11]. A linear phase diffraction grating with a high aspect ratio was successfully produced on a 5 inch n-type $〈100〉$ silicon wafer by the photo-assisted electrochemical etching technique [28]. Because of the Talbot effect, the fringes of self-image of the phase grating can be generated at the detector plane.

Unfortunately, the period of the fringes is too small to be resolved by a CCD detector directly. Accordingly, in the general setup, an absorption grating corresponding to the periodicity and orientation of the fringe pattern of the phase grating is placed immediately in front of the scintillator as an analyzer device to form a detectable Moiré pattern. In the absorption analyzer grating setup, when using the higher X-ray photon energies, the depth of absorption grating should be increased. The high-Z materials such as Au is used to absorb the X-rays, but it cannot absorb the X-ray passing it completely especially for higher X-ray photon energy, which reduces the fringe visibility and therefore deteriorates the image quality. The structured scintillator, which integrates the analyzer grating into the scintillator, functions as both the common scintillator and the analyzer grating. It is fabricated by filling silicon pore arrays with X-ray sensitive materials such as Tl-doped cesium iodide(CsI:Tl). To allow it to function as an analyzer grating, a half period of the structured scintillator is filled with CsI:Tl, which converts X-rays into visible light. Another half period of the structured scintillator is completely made of silicon. To enable total light reflection at the silicon pore inner wall of the structured scintillator, enhancing its conversion efficiency from X-ray to visible light, silicon oxide is formed on the inner wall surface [25,29]. When X-rays penetrate through a structured scintillator, only the half pitch of filling with X-ray sensitive materials converts X-rays into the visible light that can be directly detected by a visible light CCD detector. Therefore the contrast between two half pitches in one spatial period of the structured scintillator is nearly 100% for the visible light CCD detector. Compared with an absorption grating made from Au, the structured scintillator is more suitable for broadband X-rays, especially for higher photon energy X-ray phase contrast imaging. Therefore, it can potentially provide high image quality for harder X-ray with large fields of view.

The phase stepping technique is used to retrieve the phase information [13,16]. For the$\pi$shift of the phase grating, the oscillation intensity of the specimen for each detector pixel$\left(m,n\right)$can be expressed as the Fourier series [30]

3. Experimental results

The experiment was performed using the ladder-shaped structured multi-line X-ray source and the structured scintillator. It was operated at 60 KV and a 2 mA tube current. Because of the inclination of the target, the axial length of anode was 2 mm and the total effective X-ray source size was $0.6\times 0.9m{m}^2$. It had an equivalent period of $p_0=42\mu m$ and a line width of $w_0=10.5\mu m$. The phase grating had a period of $5.6\mu m$, a duty ratio of 0.5 and a depth of $40\mu m$, which corresponded to a $\pi$ phase shift for 31 keV, the most probable photon energy generated by the multi-line X-ray source. The thickness of the structured scintillator was up to $150\mu m$, and its period was $3\mu m$ with a duty ratio of 0.5 (equal to the pitch of the phase grating's self-image at the first fractional Talbot distance). The distance between the multi-line X-ray source and the phase grating was $z_0=1.47m$, and the distance between the phase grating and the structured scintillator was $z_1=105mm$. The scintillator was directly coupled with an ANDOR $2048\times 2048$ pixels $\left(13.5\mu m/pixel\right)$ CCD camera through a fiber optical tape with a demagnification of 2. Thus, the effective scintillator pixel size is $27\mu m$, and the field of view is about $55\times 55m{m}^2$.

Figure 2 shows the experimental results of a weakly absorbing specimen, a piece of a violet leaf. The exposure time is 10 s for each original image, and the field of view is $55\times 45m{m}^2$. Figure 2(a) shows the chromo-photograph of the specimen, where a part of the footstalk and leaf apex is dried. The X-ray transmission image and the differential phase contrast image, which clearly show the contour of the leaf, are shown in Fig. 2(b) and Fig. 2(c), respectively. In the section of the dried leaves shown in Fig. 2(d) and Fig. 2(g), the small difference in the intensity of the tissues is barely discernible in the absorption images in Fig. 2(e) and Fig. 2(h), while the differential phase contrast images shown in Fig. 2(f) and Fig. 2(i) show them clearly.

 

Fig. 2 (Color online) The image of a piece of a violet leaf. (a) is the conventional digital chromo-photograph. (b) is the X-ray transmission image. (c) is the differential phase contrast image. (d), (e) and (f) are magnified sections of the chromatic, transmission and differential phase contrast images, respectively, of the dried footstalk. (g), (h) and (i) are the magnified sections of the chromatic, transmission and differential phase contrast images, respectively, of the dried leaf apex. (j) and (k) are the profile values of the transmission and differential phase contrast images, respectively, of the dried footstalk.

Download Full Size | PDF

Figure 3 shows the results obtained by our method with a biological sample containing soft and hard tissues (a chicken claw). The exposure time is 10 s for each original image, and the field of view is $55\times 55m{m}^2$. The absorption and differential phase contrast images are shown in Fig. 3(a) and Fig. 3(b), respectively. As expected, the differential phase contrast image reveals much more detail of the sample, especially for the soft tissues and cartilages, while only some highly absorbing tissues are clearly visible in transmission image. Smaller structures with higher spatial frequencies (the structures of imbricate skins, for example) are also better visualized in the phase contrast images. Finally, in the magnified images corresponding to the rectangular area shown in Fig. 3(c) and Fig. 3(d), we can also see that the phase contrast image provides more information about the nail.

 

Fig. 3 The chicken claw image. (a) is the X-ray transmission image. (b) is the differential phase contrast image. (c) and (d) are magnified sections of the transmission and differential phase contrast images, respectively, of the chicken toes.

Download Full Size | PDF

4. Conclusion

On the basis of the key improvements described here, the non-absorption grating X-ray phase contrast imaging technique presented in this work represents an important progress towards applying X-ray phase contrast imaging. As a powerful inspection tool for materials composed of low-Z elements, this technique may play important roles in soft tissue pathologies, biology, paleontology and materials science. Due to the characteristics of the ladder-shaped multi-line X-ray source and the structured scintillator, the approach creates a great number of preponderant applications for higher X-ray photon energies. We believe that our method significantly advances this technique toward future practical applications.

Grating-based X-ray phase contrast imaging has long been restricted by the use of absorption gratings. Unlike absorption gratings, the ladder-shaped multi-line X-ray source and large-area structured scintillator are inexpensively fabricated in ordinary laboratory, which provides another choice for X-ray phase contrast imaging and potentially promote the use of this technique for higher X-ray photon energies. Furthermore, like other techniques, implementing X-ray dark field imaging [30] and X-ray phase contrast tomography [18], is also possible by this approach.