Meibomian glands segmentation in infrared images with limited annotation.

AIM: To investigate a pioneering framework for the segmentation of meibomian glands (MGs), using limited annotations to reduce the workload on ophthalmologists and enhance the efficiency of clinical diagnosis. METHODS: Totally 203 infrared meibomian images from 138 patients with dry eye disease, accompanied by corresponding annotations, were gathered for the study. A rectified scribble-supervised gland segmentation (RSSGS) model, incorporating temporal ensemble prediction, uncertainty estimation, and a transformation equivariance constraint, was introduced to address constraints imposed by limited supervision information inherent in scribble annotations. The viability and efficacy of the proposed model were assessed based on accuracy, intersection over union (IoU), and dice coefficient. RESULTS: Using manual labels as the gold standard, RSSGS demonstrated outcomes with an accuracy of 93.54%, a dice coefficient of 78.02%, and an IoU of 64.18%. Notably, these performance metrics exceed the current weakly supervised state-of-the-art methods by 0.76%, 2.06%, and 2.69%, respectively. Furthermore, despite achieving a substantial 80% reduction in annotation costs, it only lags behind fully annotated methods by 0.72%, 1.51%, and 2.04%. CONCLUSION: An innovative automatic segmentation model is developed for MGs in infrared eyelid images, using scribble annotation for training. This model maintains an exceptionally high level of segmentation accuracy while substantially reducing training costs. It holds substantial utility for calculating clinical parameters, thereby greatly enhancing the diagnostic efficiency of ophthalmologists in evaluating meibomian gland dysfunction.

Measures of carbohydrate quality and their association with diet quality and cardiometabolic health outcomes in Singapore middle-aged and older adults.

BACKGROUND AND AIMS: Carbohydrate quality may play a key role in cardiometabolic health and disease risk. This study aimed to assess the dietary carbohydrate quality of the free-living middle-aged and older adults in Singapore, and its association with overall diet quality and cardiometabolic health. METHODS AND RESULTS: This cross-sectional study examined the diet and cardiometabolic disease risk indicators of middle-aged and older adults in Singapore (n = 104). Dietary carbohydrate quality was assessed as the pass and fail rate of the population to four measures of carbohydrate quality: (i) dietary fiber recommended daily allowance (RDA), (ii) whole-grain recommendation, (iii) free sugar recommendation, and (iv) carbohydrate metrics. The association between each carbohydrate quality measure and diet quality, as well as cardiometabolic health, was assessed. Except for free sugar recommendation, the carbohydrate quality of the population was found to be poor with a low adherence (20-36%) to three measures. Subjects meeting these measures had generally higher intakes of fiber, protein, and most micronutrients compared with subjects who failed. Meeting different variants of the carbohydrate metrics was associated with 60% lower odds of pre-hypertensive blood pressure (p = 0.037; p = 0.047), and meeting the dietary fiber RDA was associated with lower waist circumference (p = 0.021). CONCLUSION: An improvement in carbohydrate quality is warranted among free-living middle-aged and older adults in Singapore. Not all measures of carbohydrate quality were equally effective in preserving overall diet quality; the carbohydrate metrics and dietary fiber RDA can be identified as effective measures in relation to cardiometabolic disease risk. CLINICAL TRIAL REGISTRATION: https://clinicaltrials.gov/ CLINICAL TRIAL REGISTRATION: NCT03554954, 13 Sept. 2018.

Rapid detection of the New Delhi metallo-ß-lactamase (NDM) gene by recombinase polymerase amplification.

New Delhi metallo-ß-lactamase (NDM) is a series of enzyme conferring resistance to ß-lactam antibiotics including the carbapenems. The blaNDM gene has been reported in a variety of Gram-negative bacilli, especially in the Enterobacteriaceae and Acinetobacter spp., which is deeply disconcerting for public health worldwide. In this study, recombinase polymerase amplification assays using a basic detection (Basic-RPA) and a real-time fluorescent detection (Exo-RPA) were established for detecting blaNDM gene. The RPA reactions were performed at 39 °C and finished within 20 min. Using different copy numbers of pMD18T-NDM plasmid DNA as templates, we identified the detection limit of Basic-RPA assay (1.85 × 103 copies/µL), conventional PCR assay (1.85 × 104 copies/µL), Exo-RPA assay (1.85 × 102 copies/µL) and real-time PCR assay (1.85 × 102 copies/µL). Both Basic-RPA and Exo-RPA assays were highly specific for detecting blaNDM, as there were no cross-reactions with the strains without blaNDM gene. Examination of 62 clinical samples by RPA assays and PCR assays showed the same results, suggesting that RPA assays are reliable in clinical diagnosis. The amplification time of RPA is much shorter than that of other molecular techniques, it is easy to implement and has the potential for clinical application.