Based on multimodal measurement methods of NASA task load index (NASA-TLX), task performance, surface electromyography (sEMG), heart rate (HR), and functional near-infrared spectroscopy (fNIRS), this study conducted experimental measurements and analyses under 16 different load levels of physical fatigue and mental fatigue combination conditions. This study observed the interaction between physical fatigue and mental fatigue at different levels, and at the subjective level, the effect of physical fatigue on mental fatigue was greater than that of mental fatigue on physical fatigue. Secondly, the results of fNIRS analysis showed that the premotor cortex is affected by physical fatigue, and the dorsolateral prefrontal cortex is affected by mental fatigue. Finally, this study constructed a fatigue classification model with an accuracy of 95.3%, which takes multimodal physiological data as input and 16 fatigue states as output. The research results will provide a basis for fatigue analysis, evaluation, and improvement in complex working situations.
BackgroundThe aim of the present study was to estimate the incidence, years lived with disability (YLDs), and cause of eye injury at global, regional, and national levels by age and sex based on the Global Burden of Diseases, Injuries, and Risk Factors Study (GBD) 2019.MethodsThis is a retrospective demographic analysis based on aggregated data. GBD 2019 included the burden of eye injury worldwide and its temporal and spatial characteristics in the past three decades. The Bayesian meta-regression tool and DisMod-MR 2.1 were used to analyse the estimates based on a linear regression mode of the age-standardised rates (ASR). Average annual percent change (AAPC) was calculated to represent the temporal trends of the ASR.FindingsGlobally, there were 59,933.29 thousand (95% uncertainty interval [UI]: 45,772.34–77,084.03) incident cases and 438.4 thousand (95% UI: 132.44–898.38) YLDs of eye injury in 2019. Both the ASR of incidence and YLDs decreased from 1990 to 2019, with AAPC −0.46 (95% confidence interval [CI]: −0.52 to −0.39) and −0.45 (95% CI: −0.52 to −0.39), respectively. Males had higher rates of incidence and YLDs in all age groups. Young and middle-aged adults had higher disease burdens. Regionally, Australasia had the highest ASR of YLDs to be 9.51 (95% UI: 3.00–19.58) per 100,000. Nationally, New Zealand had the highest burden of eye injury to be 11.33 (95% UI: 3.57–23.10) per 100,000. Foreign bodies, exposure to mechanical forces, and falls were the main causes of global eye injury burden in 2019, and there was an increased worldwide burden due to road injuries and executions and police conflict compared with 1990.InterpretationOur findings suggest that the incidence and burden of eye injury have decreased over the last 30 years, while the absolute number of eye injuries has substantially increased, representing a major public health concern. Males and young adults were affected to a greater degree than females and elder individuals. More attention should be paid to road injuries and executions and police conflict in order to prevent eye injury.FundingGuangdong Provincial People's Hospital (GDPH) Supporting Fund for Talent Program (KY0120220263).
BACKGROUND: The performance of healthcare workers directly impacts patient safety and treatment outcomes. This was particularly evident during the coronavirus disease 2019 (COVID-19) pandemic. OBJECTIVE: This study aimed to analyze research trends on factors influencing work performance among healthcare workers through bibliometric analysis and conduct a comparative analysis from macro and micro perspectives before and after the COVID-19 pandemic to complement the existing research. METHODS: This study involved a bibliometric analysis of 1408 articles related to work performance in the healthcare field published between 2010 and 2023, using the Web of Science, Scopus, and PubMed databases, and 37 articles were selected to determine the factors influencing work performance. RESULTS: By conducting a bibliometric analysis of the articles based on country, institution, journal, co-cited references, and keywords, this study identified a significant growth trend regarding the factors influencing work performance in the healthcare field, and research hotspots shifted from organizational factors like standard towards psychological factors such as burnout, anxiety, and depression following the outbreak of the COVID-19 pandemic. Subsequently, this study extracted 10 micro-level and 9 macro-level influencing factors from the selected articles for supplementary analysis. Furthermore, this study conducted a comparative analysis of the impact of these factors on work performance before and after the COVID-19 pandemic. CONCLUSIONS: This study addressed the limitations of previous studies regarding incomplete extraction of factors influencing work performance and unclear comparisons of parameters before and after the COVID-19 pandemic. The findings provide insights and guidance for improving the performance of healthcare workers.
Early diagnosis of pulmonary nodules is crucial for preventing and treating pulmonary cancer. The diverse and complex characteristics of pulmonary nodules, such as shape, size, speculation, and texture, pose challenges in clinical diagnosis, which can be aided by computer-aided diagnosis (CAD). However, different deep learning methods exhibit varying performances in CAD, and the lack of model interpretability hinders doctors' understanding of CAD results. To address these challenges, we propose a multitask Swin Transformer (MTST) model based on the Swin Transformer for feature extraction, which contains a multitask layer that can simultaneously output benign and malignant binary classification, multilevel classification, and pulmonary nodule features. Furthermore, we use a GAN model based on the U-Net structure (U-Net GAN)-generated image augmentation for training the network. Experimental results based on the LIDC-IDRI dataset show that, compared to the current commonly used CNN networks, our proposed MTST achieves better performance in multiple tasks, such as benign and malignant binary classification, multilevel classification, and nodule feature evaluation, and is more in line with clinical practice requirements.