• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2021.10a
• /
• pp.183-185
• /
• 2021

Detecting bacteria in blood could be an important research area in medicine and computer vision. In this paper, we propose a method for detecting bacteria in blood from 366 darkfield microscopy images acquired at Kaggle. Generate a training dataset through preprocessing and data augmentation using image processing techniques, and define a deep learning model for learning it. As a result of the experiment, it was confirmed that the proposed deep learning model effectively detects red blood cells and bacteria in darkfield microscopy images. In this paper, we learned using a relatively simple model, but it seems that more accurate results can be obtained by using a deeper model.

• Journal of Periodontal and Implant Science
• /
• v.33 no.2
• /
• pp.247-257
• /
• 2003

The purpose of this study was to evaluate the difference of subgingival bacterial compositions between periodontally healthy and diseased sites. Subgingival plaque samples were obtained from 100 sites in 20 untreated adult periodontitis patients(experimental group), and 100 sites in healthy individuals(contro1 group). Before sampling, probing pocket depth(PPD) and clinical level of attachment(CAL), Plaque Index(PI), and Sulcus Bleeding Index(SBI) were recorded for each sampled sites. Microbial samples were collected from the bases of gingival sulci or periodontal pockets with sterile curettes. The samples were examined under darkfield microscope(${\times}$400). At least 150 bacteria were evaluated and categorized on the basis of bacterial morphology and motility, i.e. cocci, non-motile rods, motile rods, and spirochetes. In control group, subgingival microbial flora consisted of 73.7% of cocci, 20.0% of non-motile rods, 4.3% of motile rods, and 2.0% of spirochetes. The microbial samples from experimental group consisted of 51.5% of cocci, 19.4% of non-motile rods, 17.6% of motile rods, and 11.6% of spirochetes. The proportion of cocci was higher in control group than in experimental group. Proportions of motile rods and spirochetes were higher in experimental group than in control group. The proportion of nonmotile rods in experimental group and control group was not significantly different. Sulcus Bleeding Index and Plaque Index showed high correlation with the bacterial composition. These findings suggests that examination of subgingival bacterial proportion may serve as more sensitive mirror of the local periodontal status than clinical parameters.