Comparison of deep learning-assisted blinking analysis system and Lipiview interferometer in dry eye patients: a cross-sectional study.

BACKGROUND: Abnormal blinking pattern is associated with ocular surface diseases. However, blink is difficult to analyze due to the rapid movement of eyelids. Deep learning machine (DLM) has been proposed as an optional tool for blinking analysis, but its clinical practicability still needs to be proven. Therefore, the study aims to compare the DLM-assisted Keratograph 5M (K5M) as a novel method with the currently available Lipiview in the clinic and assess whether blinking parameters can be applied in the diagnosis of dry eye disease (DED). METHODS: Thirty-five DED participants and 35 normal subjects were recruited in this cross-sectional study. DED questionnaire and ocular surface signs were evaluated. Blinking parameters including number of blinks, number of incomplete blinking (IB), and IB rate were collected from the blinking videos recorded by the K5M and Lipiview. Blinking parameters were individually collected from the DLM analyzed K5M videos and Lipiview generated results. The agreement and consistency of blinking parameters were compared between the two devices. The association of blinking parameters to DED symptoms and signs were evaluated via heatmap. RESULTS: In total, 140 eyes of 70 participants were included in this study. Lipiview presented a higher number of IB and IB rate than those from DLM-assisted K5M (P ≤ 0.006). DLM-assisted K5M captured significant differences in number of blinks, number of IB and IB rate between DED and normal subjects (P ≤ 0.035). In all three parameters, DLM-assisted K5M also showed a better consistency in repeated measurements than Lipiview with higher intraclass correlation coefficients (number of blinks: 0.841 versus 0.665; number of IB: 0.750 versus 0.564; IB rate: 0.633 versus 0.589). More correlations between blinking parameters and DED symptoms and signs were found by DLM-assisted K5M. Moreover, the receiver operating characteristic analysis showed the number of IB from K5M exhibiting the highest area under curve of 0.773. CONCLUSIONS: DLM-assisted K5M is a useful tool to analyze blinking videos and detect abnormal blinking patterns, especially in distinguishing DED patients from normal subjects. Large sample investigations are therefore warranted to assess its clinical utility before implementation.

Impacts of gender and age on meibomian gland in aged people using artificial intelligence.

Purpose: To evaluate the effects of age and gender on meibomian gland (MG) parameters and the associations among MG parameters in aged people using a deep-learning based artificial intelligence (AI). Methods: A total of 119 subjects aged ≥60 were enrolled. Subjects completed an ocular surface disease index (OSDI) questionnaire, received ocular surface examinations including Meibography images captured by Keratograph 5M, diagnosis of meibomian gland dysfunction (MGD) and assessment of lid margin and meibum. Images were analyzed using an AI system to evaluate the MG area, density, number, height, width and tortuosity. Results: The mean age of the subjects was 71.61 ± 7.36 years. The prevalence of severe MGD and meibomian gland loss (MGL) increased with age, as well as the lid margin abnormities. Gender differences of MG morphological parameters were most significant in subjects less than 70 years old. The MG morphological parameters detected by AI system had strong relationship with the traditional manual evaluation of MGL and lid margin parameters. Lid margin abnormities were significantly correlated with MG height and MGL. OSDI was related to MGL, MG area, MG height, plugging and lipid extrusion test (LET). Male subjects, especially the ones who smoke or drink, had severe lid margin abnormities, and significantly decreased MG number, height, and area than the females. Conclusion: The AI system is a reliable and high-efficient method for evaluating MG morphology and function. MG morphological abnormities developed with age and were worse in the aging males, and smoking and drinking were risk factors.

Determination and pharmacokinetic study of echinatin by UPLC-MS/MS in rat plasma.

Echinatin, one of the bioactive components of licorice, has exhibited diverse therapeutic effects, including anti-inflammatory and anti-oxidant effects. However, determination and pharmacokinetic study of echinatin in biomatrices have not been conducted. In this study, a simple and fast ultra performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS) method for the quantification of echinatin in rat plasma was developed, fully validated and subsequently well used in a pharmacokinetic research of echinatin after oral and intravenous administration. Rat plasma samples were operated with a simple one-step acetonitrile precipitation, and licochalcone A was used as the internal standard. Chromatographic separation of echinatin was conducted using an UPLC BEN C18 column and a gradient water (containing 0.1% formic acid)-acetonitrile mobile phase. A Waters XEVO TQS-micro Triple-Quadrupole Tandem Mass Spectrometer operating in positive electrospray ionization mode was used for detection. The approach was proved to be linear in the range of 1-1000 ng/mL and well satisfy the requirements from the guidelines of FDA. A pharmacokinetic study of echinatin was carried out by the new developed method following intravenous and oral administration to adult male Sprague-Dawley rats. Echinatin was demonstrated to be quickly absorbed and eliminated and extensively distributed with an absolute bioavailability of approximately 6.81%.

Metabolomics Analysis in Acute Paraquat Poisoning Patients Based on UPLC-Q-TOF-MS and Machine Learning Approach.

Most paraquat (PQ) poisoned patients died from acute multiple organ failure (MOF) such as lung, kidney, and heart. However, the exact mechanism of intoxication is still unclear. In order to find out the initial toxic mechanism of PQ poisoning, a blood metabolomics study based on ultraperformance liquid chromatography coupled to quadrupole time-of-flight mass spectrometry (UPLC-Q-TOF-MS) and efficient machine learning approach was performed on 23 PQ poisoned patients and 29 healthy subjects. The initial PQ plasma concentrations of PQ poisoned patients were >1000 ng/mL, and the blood samples were collected at before first hemoperfusion (HP), after first HP, and after last HP. The results showed that PQ poisoned patients all differed from healthy subjects, whatever they were before or after first HP or after last HP. The efficient machine learning approaches selected key metabolites from three UPLC/Q-TOF-MS data sets which had the highest classification performance in terms of classification accuracy, Matthews Correlation Coefficients, sensitivity, and specificity, respectively. The mass identification revealed that the most important metabolite was adenosine, which sustained in low level, regardless of whether PQ poisoned patients received HP treatment. In conclusion, decreased adenosine was the most important metabolite in PQ poisoned patients. The metabolic disturbance caused by PQ poisoning cannot be improved by HP treatment even the PQ was cleared from the blood.