Cellular Imaging Analysis Algorithm-Based Assessment and Prediction of Disease in Patients with Acute Lung Injury. (22nd August 2022)

• Record Type: Journal Article
• Authors: Gao, Liang; Xiao, Chengwang; Cheng, Taoyi; Wang, Zhaohan; Xia, Wenhan
• Other Names: Teekaraman Yuvaraja Academic Editor.
• Is Part Of: Contrast media & molecular imaging. Volume 2022(2022)
• Publication Date: 2022-08-22
• Subjects: Diagnostic imaging -- Periodicals; Magnetic resonance imaging -- Periodicals; Contrast media (Diagnostic imaging) -- Periodicals; Contrast Media -- Periodicals; Diagnostic Imaging -- Periodicals; Substances de contraste -- Périodiques; Diagnostics moléculaires -- Périodiques; Imagerie médicale; Substance de contraste; Périodique électronique (Descripteur de forme); Ressource Internet (Descripteur de forme); 616.0754
• DOI: 10.1155/2022/3193671
• Languages: English
• ISSNs: 1555-4309

Abstract

Abstract : This paper uses cellular imaging analysis algorithms to assess and predict the condition of patients with acute lung injury. Given the unique optical properties of UCNPs, this paper designs a ratiometric upconversion fluorescent nanoprobe for the determination of nitric oxide (NO) content in living cells and tissues. To address the image degradation phenomenon of optical sections, this paper uses a blind deconvolution method to abate the degradation effect caused by the scattered focus surface, thus completing the image recovery. After that, grayscale and binarization are performed using the weighted average method and the Otsu method. In this paper, we propose a migration learning-based Resnet-50 network for the triple classification of unlabeled leukocytes based on the characteristics of cell images acquired by a miniaturized label-free microfluidic cell imaging detection device. The migration learning can rapidly optimize the network parameters, the short connection structure of Resnet-50 is more suitable for feature extraction of unlabeled leukocytes than the InceptionV3 model without a short connection structure, and the accuracy of the Resnet-50 network can reach 94% in the test set. In this paper, we propose two tracking algorithms based on the dynamic Gaussian mixture model and mathematical morphology-based algorithms suitable for cells of different shapes for cell tracking in microscopic images, neuronal cell labeling in fluorescent images, and cell segmentation in mice. These methods have the advantages of low cost, speed, reproducibility, and objectivity, and we hope that their elicitation will be useful for relevant cell biology research.