research papers
Fast digital lossy compression for Xray ptychographic data
^{a}Department of Physics and Astronomy, Northwestern University, Evanston, IL 60208, USA, ^{b}Advanced Photon Source, Argonne National Laboratory, Argonne, IL 60439, USA, and ^{c}Chemistry of Life Processes Institute, Northwestern University, Evanston, IL 60208, USA
^{*}Correspondence email: cjacobsen@anl.gov
Increases in Xray
from synchrotron light sources lead to a requirement for higher frame rates from hybrid pixel array detectors (HPADs), while also favoring charge integration over However, transfer of the full uncompressed data will begin to constrain detector design, as well as limit the achievable continuous frame rate. Here a data compression scheme that is easy to implement in a HPAD's applicationspecific integrated circuit (ASIC) is described, and how different degrees of compression affect image quality in ptychography, a commonly employed coherent imaging method, is examined. Using adaptive encoding quantization, it is shown in simulations that one can digitize signals up to 16383 photons per pixel (corresponding to 14 bits of information) using only 8 or 9 bits for data transfer, with negligible effect on the reconstructed image.Keywords: Xray ptychography; pixel array detectors; lossy compression.
1. Introduction
The et al., 2014). Many light source experiments require hybrid pixel array detectors (HPADs), where a semiconductor sensor is paired with an applicationspecific integrated circuit (ASIC) to record a twodimensional intensity distribution. These HPADs can be used for direct image recording when sufficient geometric image magnification is used (Vagovič et al., 2013; Blackhall et al., 2014), but they are more commonly used to record farfield diffraction intensities in crystallography, in photon correlation spectroscopy, and in coherent diffraction imaging methods such as ptychography.
of synchrotron light sources for Xray experiments has been increasing dramatically over the past decades, with diffractionlimited storage rings beginning to deliver the next advance (ErikssonWhen an Xray photon of energy E is absorbed in an HPAD's sensor, an electron–hole separation charge q = E/W is generated, where W = 3.65 eV in the case of a silicon sensor (Fraser et al., 1994). In a photoncounting detector, this charge is collected and a photon is counted by the ASIC when this charge exceeds a threshold value which is some fraction of q. However, it takes a finite time to collect this charge due to the transport properties of the sensor, leading to a `deadtime' t_{dead} before another photon can be successfully detected. For this reason, photondetecting HPADs usually have a perpixel count rate limit of about 10^{6} photons s^{−1} (Trueb et al., 2012). This is begining to limit their application with increasingly bright Xray sources, where many photons arrive within specific time intervals due to the electron bunch structure in the storage ring. In contrast, chargeintegrating detectors become favored as Xray increases because they do not have an intrinsic limit to photon arrival rate, even in the case of Xray freeelectron lasers (XFELs) where all the photons might arrive within 20 fs. In these detectors, a total charge Q = nq is collected from n photons during an acquisition time t_{frame}, leading to a voltage V = Q/C over a collection capacitance C. This voltage then leads to an analog detection unit (ADU) of ADU = aV, where a indicates the calibration of an analogtodigital converter (ADC). Thus, in the end, the digitized signal per acquisition time t_{frame} is given by
In the following, we will assume that aE/(WC) = 1 for simplification.
In a chargeintegrating HPAD, one must periodically integrate the collected charge Q and either store that information on the ASIC or transfer it off the detector immediately (Graafsma et al., 2016). Some HPADs developed for XFEL applications include perpixel capacitors on the ASIC to store the charge Q for up to eight frames at a burst frame rate up to 10 MHz (Philipp et al., 2016), or 352 frames at a burst frame rate of 6.5 MHz (Henrich et al., 2011), followed by digitization of the charge on each capacitor and subsequent digital transfer of the detected frames. However, the maximum continuous frame rate in these detectors is no higher than 16 kHz (Allahgholi et al., 2019), or 20 kHz as anticipated by nearterm extension of the ePix detectors (Blaj et al., 2016). The common bottleneck limiting frame rate in chargeintegrating HPADs is data bandwidth (Graafsma et al., 2016), and it can limit bandwidth in photoncounting detectors as well. Some detectors such as the EIGER can be configured to store data in different bit depths (Dinapoli et al., 2011). At low incident fluence into all detector pixels, one can switch to a mode with a lower bit depth to increase the frame rate. However, this does not solve the problem at high fluence, or for situations where some pixels (for example, near the center of coherent diffraction patterns) see high fluence while others do not. Another way to reduce data bandwidth requirements is to design the ASIC with perpixel analogtodigital conversion followed by lossy compression. Immediate digitization also has the advantage of reducing signal distortion, analog noise addition, and interference between pixels. However, one must then consider how lossy compression affects the information obtained from an experiment. This is what we consider for the case of Xray ptychographic imaging.
2. Xray ptychography
Ptychography is an imaging method where a spatially limited coherent beam (the probe) illuminates a series of overlapping positions on an extended specimen, with farfield diffraction patterns recorded at each probe position (Hoppe, 1969a,b). An iterative algorithm (Faulkner & Rodenburg, 2004) is then used to recover the phase in the set of diffraction patterns, and reconstruct the magnitude and phase of the exit wave leaving the specimen with a spatial resolution limited not by the size of the probe beam but by the largest diffraction angle at which significant scattering is recorded and phased. Following its first demonstration in Xray imaging (Rodenburg et al., 2007), Xray ptychography has been adopted widely, achieving sub10 nm spatial resolution (Shapiro et al., 2014) and being used for 3D imaging via ptychographic tomography (Dierolf et al., 2010a). Reconstruction algorithms have been extended (Thibault & Menzel, 2013) to allow the probe to be scanned rapidly across the specimen (Pelz et al., 2014). However, the frame rate of currently available Xray HPADs sets a limit to highthroughput imaging demonstrations (Deng et al., 2019), so that further advances in throughput will require frame rates well above what is currently available (Jacobsen et al., 2017).
It is desirable to limit the N photons per detector pixel, one will have fluctuations between different measurements of the same intensity due to Poisson statistics with a standard deviation of . The signaltonoise ratio is then expected to be proportional to = . A number of studies have addressed the performance of iterative phase retrieval algorithms at these low photon exposures (Huang et al., 2009; Schropp & Schroer, 2010; Godard et al., 2012; Jahn et al., 2017; Hagemann & Salditt, 2017; Du et al., 2020). These studies have generally shown that the achievable spatial resolution is consistent with what one would expect based on the fluence, and the specimen's intrinsic contrast. In addition, no scatteringanglelimiting and inefficient optics are placed between the specimen and the detector in ptychography, again minimizing the radiation dose associated with imaging at a given spatial resolution and specimen contrast.
(cumulative number of photons incident per area on the specimen) both to speed up imaging time and also to minimize the Xray radiation dose deposited in the specimen. At expected signals of3. Ptychography with lossy compressed data
Our goal is to understand how varying degrees of lossy data compression affect the quality of ptychographic image reconstructions. While our work is motivated by an interest in incorporating lossy data compression on the ASIC of an HPAD, lossy data compression has been demonstrated in Xray ptychography for two different purposes: decreasing the disk storage space of a dataset (Loetgering et al., 2017), and increasing the number of diffraction patterns (detector frames) that can be processed within the memory limit of a graphical processing unit (GPU) (Wakonig et al., 2020). In the first example of decreasing disk storage space, two methods were tested: the use of singular value decomposition compression (SVDC) on the set of obtained diffraction patterns, and storing sums from noncontiguous pixels in an approach called constrained pixel sum compression (CPSC). These approaches showed a slight loss of spatial resolution when lossy SVDC was used, and a larger decrease in spatial resolution when CPSC was used (Loetgering et al., 2017). In the second case, the software package PtychoShelves (Wakonig et al., 2020) uses a scheme where the actual value N_{i} for the signal at pixel i is encoded via a quantization step (QS) into a compressed value M_{i} of
so as to use the Gaussian scaling that approximates Poisson statistics. With QS = 0.5, this means that N_{i} = 16320 is stored as M_{i} = 255, compressing nearly 14 bits of raw signal into an 8 bit integer. One can then go from the encoded value M_{i} back to a decoded value using
which does not exactly reproduce the results of Fig. 10 of Wakonig et al. (2020), though the results of that figure and of equation (3) are both within of the correct value N_{i} so that they both reproduce the signal within one standard deviation of the To test the effects of lossy compression, they simulated the generation of a ptychographic dataset with Poissondistributed signaldependent noise, compressed it with varying values of QS, and evaluated the resulting reconstructed image using two metrics. The first was the signaltonoise ratio calculated using the ground truth object; this gave a signaltonoise ratio that stayed within 5% of the uncompressed value with QS 0.5, and within 20% of the uncompressed value with QS 1.0. The second was the achieved spatial resolution of the reconstructed image. They calculated the Fourier ring correlation (FRC) as a function of scattering angle to judge the spatial resolution based on a halfbit FRC threshold criterion (van Heel & Schatz, 2005). They found essentially no change in the spatial resolution of the reconstructed image with values of QS 1.9.
We wished to consider lossy compression schemes that can be implemented on a perpixel basis. By doing so, one reduces the volume of data that must be channeled per pixel to the data output region of the ASIC in an HPAD, both increasing the aggregate data bandwidth of the ASIC and also reducing wiring demands on the ASIC layout. Even if one uses further compression downstream of the individual pixel level within the ASIC, or offASIC compression such as in a fieldprogrammable gate array (FPGA), compression at the origin of perpixel data will increase the throughput of such subsequent data handling elements (Hammer et al., 2020). While the lossy compression scheme of equation (2) can be written with a simple mathematical formula, it is less straightforward to implement on a perpixel basis in an ASIC due to the use of square root and division operations. [One example (Suresh et al., 2013) of an onchip implementation of a floating point square root calculation involves an area of 118 µm^{2} in a 65 nm node process, which is comparable with the entire area of one pixel in many HPADs and thus clearly impractical to implement on a perpixel basis.] The method of lossy SVDC is even more difficult to implement on the ASIC, in part because it requires access to the entire dataset before compression can take place. The constrained pixel sum compression (CPSC) method is more amenable to implementation on a HPAD ASIC, but it showed a more significant degradation in spatial resolution.
We therefore consider an alternative lossy compression scheme that recognizes the intrinsic signaldependent Poisson error as equation (2) does, but which is easier to incorporate in simple perpixel ASIC circuitry. It involves three steps that are easy to carry out in integer mathematics: (i) comparison with a 2^{integer} boundary value to determine an integer range value r, (ii) division by a 2^{integer} number D_{r} which is a bitwise shift unique to each count region as shown in Fig. 1, and (iii) an addition of a 2^{integer} offset number F_{r}. The formula for lossy encoding an actual count N_{i} to an encoded value M_{i} is
where the floor(x,1) function truncates any noninteger result of the division of x, as is the case with division implemented as a bitwise shift. The formula for decoding M_{i} to a lossy recovered value is
where multiplication by D_{r} can again be done using a bitwise shift (though since this decoding will be done on the computer that processes the data, rather than on the detector ASIC that encodes the data, this multiplication can alternatively be done as a floating point operation). This encoding logic is expected to occupy a 25 µm × 25 µm space for each pixel with pixel size of around 100 µm in 65 nm technology (Hammer et al., 2020). An example of an implementation of this approach is shown in Table 1, where the values of D_{r} and F_{r} used are those of scheme (a) shown in Fig. 1. As can be seen, this produces gaps in the sequence of M_{i}, but it allows a 14 bit integer (2^{14}1 = 16383) to be stored within a 9 bit integer (2^{9}1 = 511) with an error that never exceeds one standard deviation.

In order to test the effects of increasing degrees of lossy compression in ptychographic image reconstruction, we show in Fig. 1 the above scheme (a), but also schemes (b) and (c) which provide increasing degrees of lossy data compression through larger values of D_{r}. We also show in Fig. 2 the decoded values for photon numbers N_{i} for our lossy compression schemes (a), (b), and (c), and also for the scheme of equations (2) and (3) with QS = 0.5. As can be seen, for the original counts up to 64 counts, all the encoding methods except for scheme (c) deliver decoded values that are within of N_{i} (that is, within one standard deviation of the original value N_{i}). With scheme (c), some decoded values are slightly more than one standard deviation away from the original value N_{i}.
4. Effect of lossy compression on ptychographic imaging
As noted in Section 3 above, the effects of the lossy ptychographic data compression scheme of equations (2) and (3) on ptychographic image quality have already been studied (Wakonig et al., 2020), showing that values of QS 0.5 produce almost no observable difference in image correlation, or in spatial resolution as measured by the 1/2 bit threshold of the FRC. Here we wish to carry out a similar test of the effects of the lossy compression method of equations (4) and (5) with parameters for schemes (a), (b), and (c) as indicated by Fig. 1. In order to do this, we will use as the true object a twodimensional phantom that has been designed to resemble a biological cell, and a Gaussian illumination spot (which approximates the focal spot produced by various types of Xray nanofocusing optics) as shown in Fig. 3. This phantom was developed as part of a comparison of the fluence requirements of inline Fresnel holography versus farfield coherent diffraction imaging (Hagemann & Salditt, 2017), and the same phantom has been used in a more recent comparison of the fluence requirements of both farfield and nearfield ptychography along with inline Fresnel holography (Du et al., 2020). Because Xray phase is advanced rather than retarded in materials (Larsson et al., 1924; Jacobsen, 2020), the original phantom (Hagemann & Salditt, 2017) was modified by taking its complex conjugate (Du et al., 2020). Within the 48.2% of the pixels that comprise the phantom `cell' in the entire array, the optical modulation on the incident illumination imparts a mean phase of = 0.643 rad, a singlepixel variance of = 0.037 rad, and a bound of 0 to 1 rad (this object phase contrast is representative of what one might have in soft Xray imaging; the contrast is usually lower in hard Xray imaging). Object variations with = 0.037 happen over length scales of a single pixel, while photon statistics also produce variations at the single pixel level. Therefore one can ask that the object's phase variations become greater than or equal to noise fluctuations by requiring
in this case. Prior studies (Du et al., 2020) have shown that one can then estimate the required fluence from
for this phantom's parameters. This is in fact observed in FRC crossing curves in simulation studies (using the exact same phantom) of nearfield holography [Fig. 4 of Hagemann & Salditt (2017]), and both nearfield holography and farfield ptychography [Fig. 6 of Du et al. (2020)].
In order to understand the effects of lossy compression in farfield ptychography at various fluences n_{ph}, we followed the same approach used in a previous comparison of the fluence dependence of a variety of coherent Xray imaging methods (Du et al., 2020). We assumed a finite coherent illumination spot (the probe function) with a Gaussian distribution in both magnitude and phase, using a standard deviation of 6 pixels (FWHM ≃ 14 pixels) and a phase that varied from 0 to 0.5 rad. While many researchers use spiral probe scans to avoid grid artifacts (Thibault et al., 2009; Dierolf et al., 2010b), we have modeled the use of rectangular scans which are better suited to constantvelocity continuous scanning (Pelz et al., 2014; Deng et al., 2015; Huang et al., 2015), and which do not display artifacts if sufficient probe overlap is used (Bunk et al., 2008; Huang et al., 2014). The shift between probe positions was set to 5 pixels to provide a very high degree of overlap, which is of particular importance at low fluence levels (Du et al., 2020). (Even so, a slight grid artifact is present which affects some FRC values as noted in Fig. 7.) This led to a square scan grid with 66×68 probe positions, with the probe embedded in a 72×72 pixel array. Therefore at each probe position, the appropriate 72×72 pixel subregion of the phantom was extracted as shown in Fig. 3, the phase modulation of the phantom subregion was applied, and a Fourier transform was taken to calculate the farfield diffraction intensity. We then used a random number generator with Poisson statistics to introduce noise into the diffraction pattern recorded at each probe position, with this noise based on the integrated fluence (incident photons per pixel in sample space instead of detector space) from all probe positions touching upon a single pixel. We could then choose to apply one of our encoding schemes to the set of diffraction patterns, and thus evaluate ptychographic imaging as a function of fluence for data without loss, or with our lossy compression schemes (a), (b), or (c). We show in Fig. 4 a set of diffraction patterns created in this way, over a wide range of fluences, and with no encoding versus lossy compression scheme (a).
Having obtained sets of simulated ptychographic scan data, we then carried out ptychographic image reconstruction using an approach based on automatic differentiation to calculate the gradient of the loss function using Autograd. Autograd is a Python package which differentiates standard Python, Numpy and Scipy code (Maclaurin et al., 2015). The same automatic differentiation approach was used in our previous publication (Du et al., 2020), which instead used TensorFlow as the automatic differentiation engine, but employed the same forward model and optimization algorithm.
In brief, we used a least squares (LSQ) cost function to measure the difference between the present guess f_{k} of the detected intensities based on a guess of the object, versus the `measured' intensities I_{k} of the diffraction patterns at each probe position k. Automatic differentiation was then used to guide the adjustment of the object so as to minimize the difference .
The LSQ cost function is frequently used for the optimization strategy. In fact, the commonly used algorithm ePIE (Maiden & Rodenburg, 2009) is a special form of algorithms using either LSQ or Poisson cost functions (Godard et al., 2012). While reconstruction with a Poisson cost function can give sharper edge boundaries, it can also introduce fringelike artifacts around the edges of sharp features, and the formation of fringes is sensitive to initial guess. Using a Poisson cost function also has less deterministic converging behavior than using a LSQ cost function (Du et al., 2020). Moreover, the LSQ cost function was shown to have better numerical robustness in the standard deviation of the estimation (Godard et al., 2012). Therefore, for our purposes of testing the detrimental effect of lossy compression on recorded ptychographic intensities, we used the LSQ noise model for all reconstructions.
5. Numerical experiments
Using the simulated data sets and reconstruction method outlined above, we obtained reconstructed ptychographic images over a wide range of fluences n_{ph} ranging from 0.8 to 7855.8 photons per pixel assuming a beam energy of 5 keV, and with no encoding or the lossy encoding schemes (a), (b), and (c). (At the very lowest fluences, the maximum signal in any given pixel is quite low so one could use fixed 8 bit data depth with no need for compression, but we wish to demonstrate a lossy compression method that will work at the higher fluence levels required for high fidelity imaging of low contrast features.) As can be seen in Fig. 5, the use of lossy encoding scheme (a) (described in Table 1 and Fig. 1) leads to little or no reduction in image quality, even at very low fluences. However, by the time one reaches the more aggressive compression found in lossy scheme (c), there is a noticeable affect on image quality. In order to quantify this, we used two metrics. The first involves defining a finite support region S within all pixels j that contains the celllike features in the phantom. We then calculated the withinsupport mean squared error (SMSE) of the reconstructed phase of the object using
The resulting SMSE values for the images obtained using the nonencoded ptychographic data set, as well as the lossy encoded data, are shown in Fig. 6. As can be seen, there is very little change in the SMSE versus fluence when using lossy encoding schemes (a) and (b), while the more aggressive lossy encoding scheme (c) shows a significant increase in the error at all fluence values.
Another common metric for image evaluation is the Fourier ring correlation (FRC), which measures the correlation in phase of two noisy images as a function of spatial resolution (that is, as a function of spatial frequency u in the Fourier transform representation F of an image) (Saxton & Baumeister, 1982; van Heel, 1987). The maximum spatial frequency u_{max} corresponds to the Nyquist limit and is given by
where is the pixel size; the normalized spatial frequency of
with a range of 0 to 1 is used in Figs. 7 and 8. The FRC is calculated from F_{1} and the complex conjugate F_{2}^{ *} as
As noted above, an FRClike metric involving one noisy image and the ground truth image was used to measure the performance of the lossy encodingdecoding scheme of equations (2) and (3) as QS was varied (Wakonig et al., 2020). Except where noted, we used instead two noisy images from two separately generated Poissonnoiseincluded datasets, leading to image Fourier transforms F_{1} and F_{2} so that one can use equation (11) to obtain the FRC [the FRC and the FRClike metric involving the ground truth image give similar trends over changes in fluence (Du et al., 2020)]. At low spatial frequency u, two independent reconstructed images with low fluence will have high correlation due to the same rough outline of the feature existing even for poorresolution images. Since one needs more fluence to see finer features in coherent Xray imaging (Huang et al., 2009; Schropp & Schroer, 2010; Godard et al., 2012; Jahn et al., 2017; Hagemann & Salditt, 2017; Du et al., 2020), it is common to use the crossing between the FRC and a 1/2 bit threshold criterion (van Heel & Schatz, 2005) as a measure of the achieved spatial resolution in an image. For the 1/2 bit threshold, we calculated the FRC between two full images instead of two half images as described by van Heel & Schatz, (2005). The derivation of the 1/2 bit threshold formula used in this work is included in the supporting information. We show here in Fig. 8 this 1/2 bit FRC crossing point as a function of fluence for images reconstructed from the nonencoded data, and also from images reconstructed from data with lossy compression schemes (a), (b), and (c). As one approaches the critical fluence of 360 photons pixel^{−1} estimated from equation (7), the FRC crossing point nears 1 (meaning the image is reconstructed at the resolution of the pixel size) for the unencoded data, and also for lossy compression schemes (a) and (b). Again, one sees essentially no change in a metric of reconstructed image quality with lossy encoding schemes (a) and (b), but as with the SMSE error one does see a significant degradation when using lossy encoding scheme (c).
6. Conclusion
Compression of ptychographic datasets is important not only for decreasing the amount of storage space they require on disk (Loetgering et al., 2017) or in GPU memory (Wakonig et al., 2020) but it can also be used to decrease the bandwidth required for streaming high frame rate data from a hybrid pixel array detector (HPAD) on a limitedbandwidth communication channel. Compression can be provided in a FPGA attached to an ASIC, but this comes at the cost of higher power consumption (Amara et al., 2006) and furthermore it does not offer simplification in data transfer wiring within the ASIC as noted in Section 3. We have therefore described lossy compression schemes that involve simple bit shifts and additions [equations (4) and (5)], so that they can be implemented on alreadydigitized data on a perpixel basis even given the limited circuitry area available per pixel on a HPAD applicationspecific integrated circuit (ASIC). By keeping the compression loss to a value below one standard deviation of the we achieved a reduction of the data size to 64% with scheme (a) and 57% with scheme (b). With both of these schemes, the images reconstructed from simulated ptychographic data over a wide range of photon fluences show no degradation in reconstructed image quality as measured by a standard mean squared error (SMSE), and no degradation in the spatial resolution as estimated using Fourier ring correlation (FRC).
This provides a potential pathway for increasing the usable frame rate of HPADs as will be required for scaling ptychography up to imaging larger objects, and thus exploiting the high penetrating power of X rays.
Supporting information
Supplementary material. DOI: https://doi.org/10.1107/S1600577520013326/gy5014sup1.pdf
Funding information
This research used resources of the Advanced Photon Source, which is a US Department of Energy (DOE) Office of Science User Facility operated for the DOE Office of Science by Argonne National Laboratory under Contract No. DEAC0206CH11357. Funding for this research was provided by National Institute of Mental Health, National Institutes of Health (grant No. R01 MH115265).
References
Allahgholi, A., Becker, J., Delfs, A., Dinapoli, R., Göttlicher, P., Graafsma, H., Greiffenberg, D., Hirsemann, H., Jack, S., Klyuev, A., Krüger, H., Kuhn, M., Laurus, T., Marras, A., Mezza, D., Mozzanica, A., Poehlsen, J., Shefer Shalev, O., Sheviakov, I., Schmitt, B., Schwandt, J., Shi, X., Smoljanin, S., Trunk, U., Zhang, J. & Zimmer, M. (2019). Nucl. Instrum. Methods Phys. Res. A, 942, 162324. Web of Science CrossRef Google Scholar
Amara, A., Amiel, F. & Ea, T. (2006). Microelectron. J. 37, 669–677. Web of Science CrossRef Google Scholar
Blackhall, C. J., Morgan, K. S. & Pelliccia, D. (2014). Opt. Express, 22, 15437–15446. Web of Science CrossRef PubMed Google Scholar
Blaj, G., Caragiulo, P., Carini, G., Dragone, A., Haller, G., Hart, P., Hasi, J., Herbst, R., Kenney, C., Markovic, B., Nishimura, K., Pines, J., Segal, J., Tamma, C. & Tomada, A. (2016). AIP Conf. Proc. 1741, 040012. Google Scholar
Bunk, O., Dierolf, M., Kynde, S., Johnson, I., Marti, O. & Pfeiffer, F. (2008). Ultramicroscopy, 108, 481–487. Web of Science CrossRef PubMed CAS Google Scholar
Deng, J., Nashed, Y. S. G., Chen, S., Phillips, N. W., Peterka, T., Ross, R., Vogt, S., Jacobsen, C. & Vine, D. J. (2015). Opt. Express, 23, 5438–5451. Web of Science CrossRef CAS PubMed Google Scholar
Deng, J., Preissner, C. A., Klug, J. A., Mashrafi, S., Roehrig, C., Jiang, Y., Yao, Y., Wojcik, M. J., Wyman, M. D., Vine, D. J., Yue, K., Chen, S., Mooney, T., Wang, M., Feng, Z., Jin, D., Cai, Z., Lai, B. P. & Vogt, S. (2019). Rev. Sci. Instrum. 90, 083701. Web of Science CrossRef PubMed Google Scholar
Dierolf, M., Menzel, A., Thibault, P., Schneider, P., Kewish, C. M., Wepf, R., Bunk, O. & Pfeiffer, F. (2010a). Nature, 467, 436–439. Web of Science CrossRef CAS PubMed Google Scholar
Dierolf, M., Thibault, P., Menzel, A., Kewish, C. M., Jefimovs, K., Schlichting, I., König, K., Bunk, O. & Pfeiffer, F. (2010b). New J. Phys. 12, 035017. Web of Science CrossRef Google Scholar
Dinapoli, R., Bergamaschi, A., Henrich, B., Horisberger, R., Johnson, I., Mozzanica, A., Schmid, E., Schmitt, B., Schreiber, A., Shi, X. & Theidel, G. (2011). Nucl. Instrum. Methods Phys. Res. A, 650, 79–83. Web of Science CrossRef CAS Google Scholar
Du, M., Gürsoy, D. & Jacobsen, C. (2020). J. Appl. Cryst. 53, 748–759. Web of Science CrossRef CAS IUCr Journals Google Scholar
Eriksson, M., van der Veen, J. F. & Quitmann, C. (2014). J. Synchrotron Rad. 21, 837–842. Web of Science CrossRef CAS IUCr Journals Google Scholar
Faulkner, H. M. L. & Rodenburg, J. (2004). Phys. Rev. Lett. 93, 023903. Web of Science CrossRef PubMed Google Scholar
Fraser, G. W., Abbey, A. F., Holland, A., McCarthy, K., Owens, A. & Wells, A. (1994). Nucl. Instrum. Methods Phys. Res. A, 350, 368–378. CrossRef CAS Web of Science Google Scholar
Godard, P., Allain, M., Chamard, V. & Rodenburg, J. M. (2012). Opt. Express, 20, 25914–25934. Web of Science CrossRef PubMed Google Scholar
Graafsma, H., Becker, J. & Gruner, S. M. (2016). Synchrotron Light Sources and FreeElectron Lasers, edited by E. J. Jaeschke, S. Khan, J. R. Schneider & J. B. Hastings, pp. 1029–1054. Cham: Springer International Publishing. Google Scholar
Hagemann, J. & Salditt, T. (2017). J. Appl. Cryst. 50, 531–538. Web of Science CrossRef CAS IUCr Journals Google Scholar
Hammer, M., Yoshii, K. & Miceli, A. (2020). arXiv:2006.02639. Google Scholar
Heel, M. van (1987). Ultramicroscopy, 21, 95–100. Google Scholar
Heel, M. van & Schatz, M. (2005). J. Struct. Biol. 151, 250–262. Web of Science PubMed Google Scholar
Henrich, B., Becker, J., Dinapoli, R., Goettlicher, P., Graafsma, H., Hirsemann, H., Klanner, R., Krueger, H., Mazzocco, R., Mozzanica, A., Perrey, H., Potdevin, G., Schmitt, B., Shi, X., Srivastava, A. K., Trunk, U. & Youngman, C. (2011). Nucl. Instrum. Methods Phys. Res. A, 633, S11–S14. Web of Science CrossRef CAS Google Scholar
Hoppe, W. (1969a). Acta Cryst. A25, 495–501. CrossRef IUCr Journals Web of Science Google Scholar
Hoppe, W. (1969b). Acta Cryst. A25, 508–514. CrossRef IUCr Journals Web of Science Google Scholar
Huang, X., Lauer, K., Clark, J. N., Xu, W., Nazaretski, E., Harder, R., Robinson, I. K. & Chu, Y. S. (2015). Sci. Rep. 5, 9074. Web of Science CrossRef PubMed Google Scholar
Huang, X., Miao, H., Steinbrener, J., Nelson, J., Shapiro, D., Stewart, A., Turner, J. & Jacobsen, C. (2009). Opt. Express, 17, 13541–13553. Web of Science CrossRef PubMed Google Scholar
Huang, X., Yan, H., Harder, R. J., Hwu, Y., Robinson, I. K. & Chu, Y. S. (2014). Opt. Express, 22, 12634–12644. Web of Science CrossRef PubMed Google Scholar
Jacobsen, C. (2020). Xray Microscopy. Cambridge University Press. Google Scholar
Jacobsen, C., Deng, J. & Nashed, Y. (2017). J. Synchrotron Rad. 24, 1078–1081. Web of Science CrossRef IUCr Journals Google Scholar
Jahn, T., Wilke, R. N., Chushkin, Y. & Salditt, T. (2017). Acta Cryst. A73, 19–29. Web of Science CrossRef IUCr Journals Google Scholar
Larsson, A., Siegbahn, M. & Waller, I. (1924). Naturwissenschaften, 12, 1212–1213. CrossRef CAS Google Scholar
Loetgering, L., Rose, M., Treffer, D., Vartanyants, I. A., Rosenhahn, A. & Wilhein, T. (2017). Adv. Opt. Technol. 6, 475–483. CrossRef Google Scholar
Maclaurin, D., Duvenaud, D. & Adams, R. P. (2015). Proceedings of the 32nd International Conference on Machine Learning (ICML 2015), AutoML Workshop, 11 July 2015, Lille, France, pp. 2113–2122. Google Scholar
Maiden, A. M. & Rodenburg, J. M. (2009). Ultramicroscopy, 109, 1256–1262. Web of Science CrossRef PubMed CAS Google Scholar
Pelz, P. M., GuizarSicairos, M., Thibault, P., Johnson, I., Holler, M. & Menzel, A. (2014). Appl. Phys. Lett. 105, 251101. Web of Science CrossRef Google Scholar
Philipp, H. T., Tate, M. W., Purohit, P., Shanks, K. S., Weiss, J. T. & Gruner, S. M. (2016). J. Synchrotron Rad. 23, 395–403. Web of Science CrossRef CAS IUCr Journals Google Scholar
Rodenburg, J., Hurst, A., Cullis, A., Dobson, B., Pfeiffer, F., Bunk, O., David, C., Jefimovs, K. & Johnson, I. (2007). Phys. Rev. Lett. 98, 034801. Web of Science CrossRef PubMed Google Scholar
Saxton, W. O. & Baumeister, W. (1982). J. Microsc. 127, 127–138. CrossRef CAS PubMed Web of Science Google Scholar
Schropp, A. & Schroer, C. G. (2010). New J. Phys. 12, 035016. Web of Science CrossRef Google Scholar
Shapiro, D. A., Yu, Y.S., Tyliszczak, T., Cabana, J., Celestre, R., Chao, W., Kaznatcheev, K., Kilcoyne, A. L. D., Maia, F., Marchesini, S., Meng, Y. S., Warwick, T., Yang, L. L. & Padmore, H. A. (2014). Nat. Photon. 8, 765–769. Web of Science CrossRef CAS Google Scholar
Suresh, S., Beldianu, S. F. & Ziavras, S. G. (2013). Proceedings of the 24th IEEE International Conference on ApplicationSpecific Systems, Architectures, and Processors, 5–7 June 2013, Washington, DC, USA, pp. 269–272. IEEE. Google Scholar
Thibault, P., Dierolf, M., Bunk, O., Menzel, A. & Pfeiffer, F. (2009). Ultramicroscopy, 109, 338–343. Web of Science CrossRef PubMed CAS Google Scholar
Thibault, P. & Menzel, A. (2013). Nature, 494, 68–71. Web of Science CrossRef CAS PubMed Google Scholar
Trueb, P., Sobott, B. A., Schnyder, R., Loeliger, T., Schneebeli, M., Kobas, M., Rassool, R. P., Peake, D. J. & Broennimann, C. (2012). J. Synchrotron Rad. 19, 347–351. Web of Science CrossRef CAS IUCr Journals Google Scholar
Vagovič, P., Korytár, D., Cecilia, A., Hamann, E., Švéda, L., Pelliccia, D., Härtwig, J., Zápražný, Z., Oberta, P., Dolbnya, I., Shawney, K., Fleschig, U., Fiederle, M. & Baumbach, T. (2013). J. Synchrotron Rad. 20, 153–159. Web of Science CrossRef IUCr Journals Google Scholar
Wakonig, K., Stadler, H.C., Odstrčil, M., Tsai, E. H. R., Diaz, A., Holler, M., Usov, I., Raabe, J., Menzel, A. & GuizarSicairos, M. (2020). J. Appl. Cryst. 53, 574–586. Web of Science CrossRef CAS IUCr Journals Google Scholar
This is an openaccess article distributed under the terms of the Creative Commons Attribution (CCBY) Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original authors and source are cited.