Single Image Super-Resolution with Arbitrary Magnification Based on High-Frequency Attention Network

Among various developments in the field of computer vision, single image super-resolution of images is one of the most essential tasks. However, compared to the integer magnification model for super-resolution, research on arbitrary magnification has been overlooked. In addition, the importance of single image super-resolution at arbitrary magnification is emphasized for tasks such as object recognition and satellite image magnification. In this study, we propose a model that performs arbitrary magnification while retaining the advantages of integer magnification. The proposed model extends the integer magnification image to the target magnification in the discrete cosine transform (DCT) spectral domain. The broadening of the DCT spectral domain results in a lack of high-frequency components. To solve this problem, we propose a high-frequency attention network for arbitrary magnification so that high-frequency information can be restored. In addition, only high-frequency components are extracted from the image with a mask generated by a hyperparameter in the DCT domain. Therefore, the high-frequency components that have a substantial impact on image quality are recovered by this procedure. The proposed framework achieves the performance of an integer magnification and correctly retrieves the high-frequency components lost between the arbitrary magnifications. We experimentally validated our model’s superiority over state-of-the-art models.

Evaluation of the role of fundus fluorescein angiography and spectral domain-optical coherence tomography in choroidal neo-vascularization: a hospital-based study

Background: The study was conducted to evaluate the role of fundus fluorescein angiography (FFA) and spectral domain-optical coherence tomography (SD-OCT) in choroidal neo-vascularisation (CNV).Methods: This was a hospital based prospective study carried out in the post-graduate department of ophthalmology, SKIMS medical college, Bemina, Srinagar, Jammu and Kashmir. All patients diagnosed with CNV fulfilling the criteria during the study period w.e.f. October 2018 to March 2020 were enrolled. Visual acuity and pinhole test using Snellen’s chart for literate and E chart for illiterate patient, slit lamp biomicroscope for anterior segment examination, ophthalmoscopy, including stereoscopic examination of the posterior pole, 90D examination of the fundus, Intra-ocular pressure measurement, FFA and SD-OCT was done in these patients.Results: Diagnostic accuracy of OCT was observed with a sensitivity 79.1% (95% confidence interval (CI): 67.3-90.7), specificity 84.3% (95% CI: 74.5-92.9), positive and negative predictive values 78.7% and 85.4%, respectively, (95% CI: 65.5-95.6) and (74.8-93.4) and diagnostic accuracy of FFA was observed with a sensitivity 81.4% (95% confidence interval (CI): 70.6-93.5), specificity 82.31% (95% CI: 71.9-89.3), positive and negative predictive values 79.9% and 83.7%, respectively, (95% CI: 68.8-92.9) and (70.3-91.2).Conclusions: FFA is the gold standard procedure for screening ARMD and detection of dry ARMD, but OCT is more specific diagnostic tool in detecting early subretinal neovascular membrane and also to assess the extent, location and activity of the neovascular membranes. This study concludes that SDOCT is highly sensitive for identifying AMD, CNV, and CNV activity and due to its non-invasive nature with no adverse effects and less time consuming can be used as 1st line of diagnostic modality and FFA be reserved for cases where SD-OCT is not helpful.