High-quality computational ghost imaging under low sampling has always attracted much attention and is an important step in the practical application of computational ghost imaging. However, as far as we know, most studies focus on achieving high-quality computational ghost imaging with one single pixel detector. The high efficiency computational ghost imaging method using multiple single pixel detectors for array measurement is rarely mentioned. In this work, a new computational ghost imaging method based on deep learning technology and array detector measurement has been proposed, which can achieve fast and high-quality imaging. This method can resolve the problem of misalignment and overlap of some pixels in the reconstructed image due to the incomplete correspondence between the array detector and the light field area. At the same time, the problem of partial information loss in the reconstructed image because of the gap between the detection units of the array detector has also been solved. Simulation and experiment results show that our method can obtain high computational ghost imaging quality, even at the low sampling rate of 0.03, and as the detection unit of the array detector increases, the number of sampling is further reduced. This method improves the applicability of computational ghost imaging and can be applied to many fields such as real-time detection and biomedical imaging.

High-quality imaging under low sampling time is an important step in the practical application of computational ghost imaging (CGI). At present, the combination of CGI and deep learning has achieved ideal results. However, as far as we know, most researchers focus on one single pixel CGI based on deep learning, and the combination of array detection CGI and deep learning with higher imaging performance has not been mentioned. In this work, we propose a novel multi-task CGI detection method based on deep learning and array detector, which can directly extract target features from one-dimensional bucket detection signals at low sampling times, especially output high-quality reconstruction and image-free segmentation results at the same time. And this method can realize fast light field modulation of modulation devices such as digital micromirror device to improve the imaging efficiency by binarizing the trained floating-point spatial light field and fine-tuning the network. Meanwhile, the problem of partial information loss in the reconstructed image due to the detection unit gap in the array detector has also been solved. Simulation and experimental results show that our method can simultaneously obtain high-quality reconstructed and segmented images at sampling rate of 0.78 %. Even when the signal-to-noise ratio of the bucket signal is 15 dB, the details of the output image are still clear. This method helps to improve the applicability of CGI and can be applied to resource-constrained multi-task detection scenarios such as real-time detection, semantic segmentation, and object recognition.

AIM: To compare choroidal neovascularization (CNV) lesion measurements obtained by in vivo imaging modalities, with whole mount histological preparations stained with isolectin GS-IB4, using a murine laser-induced CNV model. METHODS: B6N.Cg-Tg(Csf1r-EGFP)1Hume/J heterozygous adult mice were subjected to laser-induced CNV and were monitored by fluorescein angiography (FA), multicolor (MC) fundus imaging and optical coherence tomography angiography (OCTA) at day 14 after CNV induction. Choroidal-retinal pigment epithelium (RPE) whole mounts were prepared at the end of the experiment and were stained with isolectin GS-IB4. CNV areas were measured in all different imaging modalities at day 14 after CNV from three independent raters and were compared to choroidal-RPE whole mounts. Intraclass correlation coefficient (ICC) type 2 (2-way random model) and its 95% confidence intervals (CI) were calculated to measure the correlation between different raters' measurements. Spearman's rank correlation coefficient (Spearman's r) was calculated for the comparison between FA, MC and OCTA data and histology data. RESULTS: FA (early and late) and MC correlates well with the CNV measurements ex vivo with FA having slightly better correlation than MC (FA early Spearman's r=0.7642, FA late Spearman's r=0.7097, and MC Spearman's r=0.7418), while the interobserver reliability was good for both techniques (FA early ICC=0.976, FA late ICC=0.964, and MC ICC=0.846). In contrast, OCTA showed a poor correlation with ex vivo measurements (Spearman's r=0.05716) and high variability between different raters (ICC=0.603). CONCLUSION: This study suggests that FA and MC imaging could be used for the evaluation of CNV areas in vivo while caution must be taken and comparison studies should be performed when OCTA is employed as a CNV monitoring tool in small rodents.

Improving imaging quality and reducing time consumption are the key problems that need to be solved in the practical application of ghost imaging. Hence, we demonstrate a double filter iterative ghost imaging method, which adopts the joint iteration of projected Landweber iterative regularization and double filtering based on block matching three dimensional filtering and guided filtering to achieve high-quality image reconstruction under low measurement and low iteration times. This method combines the advantages of ill-posed problem solution of projected Landweber iterative regularization with double filtering joint iterative de-noising and edge preservation. The numerical simulation results show that our method outperforms the comparison method by 4 to 6 dB in terms of peak signal-to-noise ratio for complex binary target ‘rice’ and grayscale target ‘aircraft’ after 1500 measurements. The comparison results of experiments and numerical simulations using similar aircraft targets show that this method is superior to the comparison method, especially in terms of richer and more accurate edge detection results. This method can simultaneously obtain high quality reconstructed image and edge feature information under low measurement and iteration times, which is of great value for the practical application fields of imaging and edge detection at the same time, such as intelligent driving, remote sensing and other fields.