Establishing and validating a spotted tongue recognition and extraction model based on multiscale convolutional neural network / 数字中医药（英文）

@#Objective In tongue diagnosis, the location, color, and distribution of spots can be used to speculate on the viscera and severity of the heat evil. This work focuses on the image analysis method of artificial intelligence (AI) to study the spotted tongue recognition of traditional Chinese medicine (TCM). Methods A model of spotted tongue recognition and extraction is designed, which is based on the principle of image deep learning and instance segmentation. This model includes multiscale feature map generation, region proposal searching, and target region recognition. Firstly, deep convolution network is used to build multiscale low- and high-abstraction feature maps after which, target candidate box generation algorithm and selection strategy are used to select high-quality target candidate regions. Finally, classification network is used for classifying target regions and calculating target region pixels. As a result, the region segmentation of spotted tongue is obtained. Under non-standard illumination conditions, various tongue images were taken by mobile phones, and experiments were conducted. Results The spotted tongue recognition achieved an area under curve (AUC) of 92.40%, an accuracy of 84.30% with a sensitivity of 88.20%, a specificity of 94.19%, a recall of 88.20%, a regional pixel accuracy index pixel accuracy (PA) of 73.00%, a mean pixel accuracy (mPA) of 73.00%, an intersection over union (IoU) of 60.00%, and a mean intersection over union (mIoU) of 56.00%. Conclusion The results of the study verify that the model is suitable for the application of the TCM tongue diagnosis system. Spotted tongue recognition via multiscale convolutional neural network (CNN) would help to improve spot classification and the accurate extraction of pixels of spot area as well as provide a practical method for intelligent tongue diagnosis of TCM.

Characterization of OAZ1 and its potential functions in goose follicular development

Background: Ornithine decarboxylase antizyme 1 (OAZ1) is an important regulator of polyamine synthesis and uptake. Our previous studies indicated that high OAZ1 expression in the ovaries of laying geese is responsible for poor egg production. In the present study, the molecular characterization of goose OAZ1 gene was analyzed, as well as the expression profile in various follicular tissues. Results: An 873-bp cDNA sequence of the OAZ1 gene (Accession No. KC845302) with a +1 frameshift site (+175T) was obtained. The sequence consisted of a 652-bp two overlapping open reading frames (a putative protein with 216 amino acids). The OAZ domain, OAZ signature and OAZ super family domain were prominent conserved regions among species. As the follicle size increased, OAZ1 abundance showed an increasing trend during follicular development, while it decreased during follicular regression. The level of OAZ1 mRNA expression was the lowest in the fifth largest preovulatory follicle, and was 0.65-fold compared to the small white follicle (P b 0.05). OAZ1 mRNA expression in the largest preovulatory and postovulatory follicle was 2.11- and 2.49-fold compared to the small white follicle, respectively (P b 0.05). Conclusions: The goose OAZ1 structure confirms that OAZ1 plays an important role in ornithine decarboxylase-mediated regulation of polyamine homeostasis. Our findings provide an evidence for a potential function of OAZ1 in follicular development, ovulation and regression.

Evidence of a role for prolactin as regulators of ovarian follicular development in goose

Background Prolactin (PRL) regulates development and reproduction, and its effects are mediated by the prolactin receptor (PRLR). In order to clarify the role of PRLR and PRL in the process of follicular development in the goose ovary, the level of PRLR mRNA expression in the ovary and follicles of the Sichuan white goose was determined, as well as the PRL concentration in ovarian follicles. Results The level of PRLR mRNA in the hierarchical follicles (HFs) initially increased, and subsequently decreased, whereas PRLR expression was initially low and later increased in postovulatory follicles (POFs). The level of PRLR mRNA expression was the highest in the F4 follicles, and lowest in the F1 follicles in all of the examined follicles. Compared with the level of PRLR mRNA expression in the small white follicles (SWFs), the level of PRLR mRNA was 2.86- and 1.44-fold higher in the F4 and small yellow follicles (SYFs), respectively (P < 0.05). The level of PRLR mRNA expression in the F4 follicles was highest (P < 0.05) in HFs. The highest PRL concentration in all of the examined samples was observed in SYFs and F1, with concentration of 6162 mLU/g and 6197 mLU/g, respectively. The PRL concentration in SYFs was significantly higher compared with SWFs (P < 0.05). Conclusions The change of PRL concentration was similar to the PRLR mRNA expression level in preovulatory follicles. These results suggest that the PRL mediated by the PRLR plays a stimulatory role in the SWF to SYF transition.