ABSTRACT
Synchrotron radiation can be used as a light source in X-ray microscopy to acquire a high-resolution image of a microscale object for tomography. However, numerous projections must be captured for a high-quality tomographic image to be reconstructed; thus, image acquisition is time consuming. Such dense imaging is not only expensive and time consuming but also results in the target receiving a large dose of radiation. To resolve these problems, sparse acquisition techniques have been proposed; however, the generated images often have many artefacts and are noisy. In this study, a deep-learning-based approach is proposed for the tomographic reconstruction of sparse-view projections that are acquired with a synchrotron light source; this approach proceeds as follows. A convolutional neural network (CNN) is used to first interpolate sparse X-ray projections and then synthesize a sufficiently large set of images to produce a sinogram. After the sinogram is constructed, a second CNN is used for error correction. In experiments, this method successfully produced high-quality tomography images from sparse-view projections for two data sets comprising Drosophila and mouse tomography images. However, the initial results for the smaller mouse data set were poor; therefore, transfer learning was used to apply the Drosophila model to the mouse data set, greatly improving the quality of the reconstructed sinogram. The method could be used to achieve high-quality tomography while reducing the radiation dose to imaging subjects and the imaging time and cost.
ABSTRACT
In this study, we evaluate the association between biometrics and intraocular pressure (IOP) during femtosecond laser-assisted cataract surgery (FLACS) in normal patients and those with open-angle glaucoma (OAG). A retrospective cross-sectional study was conducted. A total of 103 patients who had received elective FLACS were enrolled, and those with OAG who received FLACS were further divided into a subgroup. The perioperative IOP of FLACS was measured before, during, and after the suction procedure. Demographic data and preoperative biometrics were collected from the medical records. The generalized linear model was applied to yield the adjusted odds ratio (aOR) and corresponding 95% confidence interval (CI) of each biometric for the IOP elevation in the whole group and the OAG subgroup. The mean preoperative IOP was 20.96 ± 4.79 mmHg, which rose to 55.37 ± 11.58 mmHg during suction, and decreased to 23.75 ± 6.42 mmHg after suction; the IOP both during and after suction was significantly higher than the presuction IOP (both p < 0.001). The mean IOP elevation was 34.41 ± 9.70 mmHg in the whole study population, and the difference in IOP elevation between OAG and nonglaucoma subgroups was not significant (p = 0.159). In the whole group, the presuction IOP, postdilated pupil size (PPS), and central corneal thickness (CCT) were positively corrected to higher IOP elevation (all p < 0.05), while axial length (AL) was negatively related to IOP elevation (aOR: 0.020, 95% CI: 0.008-0.699, p = 0.042). For the OAG subgroup, the longer AL was more significantly correlated to lower IOP elevation compared to those without glaucoma (aOR: 0.231, 95% CI: 0.106-0.502, p = 0.006). In conclusion, presuction IOP, PPS, and CCT are related to higher IOP during FLAC, while the AL is negatively correlated to the IOP elevation in FLACS, especially for patients with OAG. Reviewing these parameters before FLACS may enable physicians to find patients who are at risk of IOP elevation.
