Search

In our research, we aimed to address the shortcomings of traditional fruit image classification models, which struggle with inconsistent lighting, complex backgrounds, and high computational demands. To overcome these challenges, we developed a novel multi-label classification method incorporating advanced image preprocessing techniques, such as Contrast Limited Adaptive Histogram Equalization and the Gray World algorithm, which enhance image quality and color balance. Utilizing lightweight encoder–decoder architectures, specifically MobileNet, DenseNet, and EfficientNet, optimized with an Asymmetric Binary Cross-Entropy Loss function, we improved model performance in handling diverse sample difficulties. Furthermore, Multi-Label Knowledge Distillation (MLKD) was implemented to transfer knowledge from large, complex teacher models to smaller, efficient student models, thereby reducing computational complexity without compromising accuracy. Experimental results on the DeepFruit dataset, which includes 21,122 images of 20 fruit categories, demonstrated that our method achieved a peak mean Average Precision (mAP) of 90.2% using EfficientNet-B3, with a computational cost of 7.9 GFLOPs. Ablation studies confirmed that the integration of image preprocessing, optimized loss functions, and knowledge distillation significantly enhances performance compared to the baseline models. This innovative method offers a practical solution for real-time fruit classification on resource-constrained devices, thereby supporting advancements in smart agriculture and the food industry. [ABSTRACT FROM AUTHOR]

Background: A lack of sleep or disorder in sleep-wake cycles has been associated with metabolic impairments. However, few studies have investigated the association between daytime napping duration and the risk of non-alcoholic fatty liver disease. This study aimed to investigate the association of daytime napping duration with the risk of non-alcoholic fatty liver disease in a Chinese population.Methods: This cross-sectional study analyzed data from the Health Management Center of Nanfang Hospital, Guangdong Province. A total of 3363 participants aged 20-79 years were recruited and admitted from January 20, 2018, to October 16, 2020. Non-alcoholic fatty liver disease was diagnosed using abdominal ultrasonography. The outcome was the association between daytime sleep duration and the risk of non-alcoholic fatty liver disease.Results: Compared with non-nappers, long daytime nappers (≥ 60 min) were associated with a higher risk of non-alcoholic fatty liver disease in the crude model (odds ratio 2.138; 95% confidence interval 1.88-2.61, P < 0.05) and in the multivariable adjustment model (odds ratio 2.211; 95% confidence interval 1.042-4.690, P < 0.05) after adjusting for demographic, educational, and metabolic risk factors. The association was moderately enhanced with additional adjustments for night sleep duration and socioeconomic or other factors (odds ratio 2.253; 95% confidence interval 1.061-4.786, P = 0.035).Conclusion: In this cross-sectional study, daytime napping duration of ≥ 60 min was positively associated with the risk of non-alcoholic fatty liver disease in an occupational population of Guangdong Province after multivariable adjustment. [ABSTRACT FROM AUTHOR]

Terahertz technology is widely used in communications, radar, electronic countermeasures, precision guidance, and biomedical applications. In this paper, a 199 to 238 GHz broadband subharmonic mixer combining two‐stage reduced matching technology with Global Design Method is designed, fabricated, measured, and analysed successfully. Compared with traditional single‐stage reduced matching technology and Subdivision Design Method, the two‐stage reduced matching technology is applied for terahertz mixer's Radio Frequency (RF) waveguide and Local Frequency (LO) waveguide, while the LO/Intermediate Frequency (IF) Low‐Pass Filter of terahertz mixers is not necessary to design separately. The core of Global Design Method is that the whole mixer circuit, including LO/RF waveguide, is divided into basic transmission units, which match and filter circuit simultaneously with more optimization variables. When LO driven power is 3.1 mW, LO frequency is 108, 112 GHz. The measured single‐sideband (SSB) conversion loss ranges from 7.84 to 12.4 dB with RF frequency 199 to 238 GHz. The measured double‐sideband (DSB) noise temperature varies between 580 and 1059 K with LO driven power 2 to 5 mW. The multi‐stage reduced matching technology and Global Design Method is general, effective, accurate, can reduce circuit size, simplify circuit structure, broaden frequency band, and can be applied for terahertz harmonically mixers. [ABSTRACT FROM AUTHOR]

Hyperuricemia (HUA) has emerged as the second most prevalent metabolic disorder characterized by prolonged and asymptomatic period, triggering gout and metabolism-related outcomes. Early detection and prognosis prediction for HUA and gout are crucial for pre-emptive interventions. Integrating genetic and clinical data from 421287 UK Biobank and 8900 Nanfang Hospital participants, a stacked multimodal machine learning model is developed and validated to synthesize its probabilities as an in-silico quantitative marker for hyperuricemia (ISHUA). The model demonstrates satisfactory performance in detecting HUA, exhibiting area under the curves (AUCs) of 0.859, 0.836, and 0.779 within the train, internal, and external test sets, respectively. ISHUA is significantly associated with gout and metabolism-related outcomes, effectively classifying individuals into low- and high-risk groups for gout in the train (AUC, 0.815) and internal test (AUC, 0.814) sets. The high-risk group shows increased susceptibility to metabolism-related outcomes, and participants with intermediate or favorable lifestyle profiles have hazard ratios of 0.75 and 0.53 for gout compared with those with unfavorable lifestyles. Similar trends are observed for other metabolism-related outcomes. The multimodal machine learning-based ISHUA marker enables personalized risk stratification for gout and metabolism-related outcomes, and it is unveiled that lifestyle changes can ameliorate these outcomes within high-risk group, providing guidance for preventive interventions., (© 2024 The Author(s). Advanced Science published by Wiley‐VCH GmbH.)

The utilization of vast amounts of EHR data is crucial to the studies in medical informatics. Physicians are medical participants who directly record clinical data into EHR with their personal expertise, making their roles essential in follow-up data utilization, which current studies have yet to recognize. This paper proposes a physician-centered perspective for EHR data utilization and emphasizes the feasibility and potentiality of digging into physicians' latent decision patterns in EHR. To support our proposal, we design a physician-centered CDS approach named PhyC and test it on a real-world EHR dataset. Experiments show that PhyC performs significantly better in the auxiliary diagnosis of multiple diseases than globally learned models. Discussions on experimental results suggest that physician-centered data utilization can help to derive more objective CDS models, while more means for utilization need further exploration.

Patients' participation plays a crucial role in the management of chronic diseases. Educating patients about their diseases allows patients to self-regulate their daily health conditions more reasonably and effectively. This study focuses on an informative way to develop daily education and guidance among patients. We provide a systematic approach to establish, process, and present the knowledge map of chronic diseases. An ontology technique is used to model clinical knowledge and rules. Rule-based inference service is constructed based on the RETE reasoning algorithm. Several considerations in semantic visualization and interaction are listed as recommendations. A prototype of Chronic Disease Knowledge Education System (CD-KES) based on this approach has been built for diabetes mellitus. With the prototype, comparative evaluations for system performances in knowledge querying and browsing are taken. The results shows the system can help patients to more easily understand medical knowledge and avoid potential negative consequences.

Clinical decision support systems (CDSSs) have been proved as an efficient way to improve health care quality. However, the inflexibility in integrating multiple clinical practice guidelines (multi-CPGs), the mass input workload of patient data, and the difficulty in system sharing become barriers of CDSSs implementation. In this paper, we proposed a framework of CDSS for chronic disease based on ontology and service-oriented architecture (SOA) to improve these defects. We used ontology for knowledge base construction on multi-CPGs integration to overcome their differences as well as reduce the input procedure of patient data by ontology reasoning. Furthermore, we built the CDSS on an SOA structure to provide flexibility in system and data sharing, such that patients could get suggestions from the same system for self-management of chronic disease. A typical case was used to validate the CDSS functions and accuracy. Two clients were developed to illustrate the SOA superiority.