Honeycombing Lung(HCL)is a chronic lung condition marked by advanced fibrosis,resulting in enlarged air spaces with thick fibrotic walls,which are visible on Computed Tomography(CT)scans.Differentiating between normal...Honeycombing Lung(HCL)is a chronic lung condition marked by advanced fibrosis,resulting in enlarged air spaces with thick fibrotic walls,which are visible on Computed Tomography(CT)scans.Differentiating between normal lung tissue,honeycombing lungs,and Ground Glass Opacity(GGO)in CT images is often challenging for radiologists and may lead to misinterpretations.Although earlier studies have proposed models to detect and classify HCL,many faced limitations such as high computational demands,lower accuracy,and difficulty distinguishing between HCL and GGO.CT images are highly effective for lung classification due to their high resolution,3D visualization,and sensitivity to tissue density variations.This study introduces Honeycombing Lungs Network(HCL Net),a novel classification algorithm inspired by ResNet50V2 and enhanced to overcome the shortcomings of previous approaches.HCL Net incorporates additional residual blocks,refined preprocessing techniques,and selective parameter tuning to improve classification performance.The dataset,sourced from the University Malaya Medical Centre(UMMC)and verified by expert radiologists,consists of CT images of normal,honeycombing,and GGO lungs.Experimental evaluations across five assessments demonstrated that HCL Net achieved an outstanding classification accuracy of approximately 99.97%.It also recorded strong performance in other metrics,achieving 93%precision,100%sensitivity,89%specificity,and an AUC-ROC score of 97%.Comparative analysis with baseline feature engineering methods confirmed the superior efficacy of HCL Net.The model significantly reduces misclassification,particularly between honeycombing and GGO lungs,enhancing diagnostic precision and reliability in lung image analysis.展开更多
This paper addresses the verification of strong currentstate opacity with respect to real-time observations generated from a discrete-event system that is modeled with time labeled Petri nets. The standard current-sta...This paper addresses the verification of strong currentstate opacity with respect to real-time observations generated from a discrete-event system that is modeled with time labeled Petri nets. The standard current-state opacity cannot completely characterize higher-level security. To ensure the higher-level security requirements of a time-dependent system, we propose a strong version of opacity known as strong current-state opacity. For any path(state-event sequence with time information)π derived from a real-time observation that ends at a secret state, the strong current-state opacity of the real-time observation signifies that there is a non-secret path with the same real-time observation as π. We propose general and non-secret state class graphs, which characterize the general and non-secret states of time-dependent systems, respectively. To capture the observable behavior of non-secret states, a non-secret observer is proposed.Finally, we develop a structure called a real-time concurrent verifier to verify the strong current-state opacity of time labeled Petri nets. This approach is efficient since the real-time concurrent verifier can be constructed by solving a certain number of linear programming problems.展开更多
This study aimed to differentiate between the therapeutic and purely cosmetic aspects of keratopigmentation by examining the results of previous studies.Corneal opacity,corneal leukoma,and iris abnormalities represent...This study aimed to differentiate between the therapeutic and purely cosmetic aspects of keratopigmentation by examining the results of previous studies.Corneal opacity,corneal leukoma,and iris abnormalities represent the primary reasons for keratopigmentation in therapeutic settings.In therapeutic contexts,the most commonly reported adverse effects included inflammation,foreign body sensation,light sensitivity,reduced intraocular pressure,and neovascularization.The side effects most frequently noted in purely cosmetic applications were photophobia and ectasia.In therapeutic scenarios,patients who experience pre-operative vision impairment often report not only the absence of visual loss post-procedure but also an enhancement in visual acuity,accompanied by alleviation of symptoms such as glare.In purely cosmetic procedures,visual acuity tends to remain stable.Research indicates that dissatisfaction with keratopigmentation results is more prevalent in studies aimed at therapeutic interventions compared to cosmetic enhancements.Furthermore,instances of discoloration and fading are more frequently observed in therapeutic cases than in cosmetic ones,including the need for retouching.The use of keratopigmentation in both therapeutic and cosmetic settings produces positive and acceptable results,especially in cosmetic applications.Nevertheless,complications are more commonly observed in therapeutic situations,which underscores the importance of candidate selection for keratopigmentation procedures.展开更多
We explored the effects of algorithmic opacity on employees’playing dumb and evasive hiding rather than rationalized hiding.We examined the mediating role of job insecurity and the moderating role of employee-AI coll...We explored the effects of algorithmic opacity on employees’playing dumb and evasive hiding rather than rationalized hiding.We examined the mediating role of job insecurity and the moderating role of employee-AI collaboration.Participants were 421 full-time employees(female=46.32%,junior employees=31.83%)from a variety of organizations and industries that interact with AI.Employees filled out data on algorithm opacity,job insecurity,knowledge hiding,employee-AI collaboration,and control variables.The results of the structural equation modeling indicated that algorithm opacity exacerbated employees’job insecurity,and job insecurity mediated between algorithm opacity and playing dumb and evasive hiding rather than rationalized hiding.The relationship between algorithmic opacity and playing dumb and evasive hiding was more positive when the level of employee-AI collaboration was higher.These findings suggest that employee-AI collaboration reinforces the indirect relationship between algorithmic opacity and playing dumb and evasive hiding.Our study contributes to research on human and AI collaboration by exploring the dark side of employee-AI collaboration.展开更多
Objective: To analyze the clinical manifestations of intraocular lens (IOL) opacity after cataract surgery through case reports, and to explore its pathogenesis and diagnosis and treatment ideas, so as to provide a ba...Objective: To analyze the clinical manifestations of intraocular lens (IOL) opacity after cataract surgery through case reports, and to explore its pathogenesis and diagnosis and treatment ideas, so as to provide a basis for the early diagnosis and correct treatment of IOL opacity. Methods: The clinical data of one patient diagnosed with IOL opacity and underwent intraocular lens replacement in the Department of Ophthalmology, Affiliated Hospital of Youjiang Medical College for Nationalities in December 2023 were reported. The characteristics of IOL opacity were observed, and the research progress and pathogenesis of IOL opacity were understood by consulting the literature. Results: This patient is the first case of IOL opacity in our hospital. The specific reason is unclear. It is considered to be related to the IOL material. Conclusion: Hydrophilic acrylic IOL is widely used in clinic because of its good histocompatibility. However, due to its hydrophilicity, there are more cases of IOL opacity than other types of IOL. At present, there is no unified conclusion on the etiology and mechanism of IOL opacity. IOL opacity can seriously affect vision and is easily misdiagnosed as a posterior cataract. We should fully disperse the large pupil and carefully observe under the slit lamp. The most effective treatment for this disease is IOL replacement.展开更多
文摘Honeycombing Lung(HCL)is a chronic lung condition marked by advanced fibrosis,resulting in enlarged air spaces with thick fibrotic walls,which are visible on Computed Tomography(CT)scans.Differentiating between normal lung tissue,honeycombing lungs,and Ground Glass Opacity(GGO)in CT images is often challenging for radiologists and may lead to misinterpretations.Although earlier studies have proposed models to detect and classify HCL,many faced limitations such as high computational demands,lower accuracy,and difficulty distinguishing between HCL and GGO.CT images are highly effective for lung classification due to their high resolution,3D visualization,and sensitivity to tissue density variations.This study introduces Honeycombing Lungs Network(HCL Net),a novel classification algorithm inspired by ResNet50V2 and enhanced to overcome the shortcomings of previous approaches.HCL Net incorporates additional residual blocks,refined preprocessing techniques,and selective parameter tuning to improve classification performance.The dataset,sourced from the University Malaya Medical Centre(UMMC)and verified by expert radiologists,consists of CT images of normal,honeycombing,and GGO lungs.Experimental evaluations across five assessments demonstrated that HCL Net achieved an outstanding classification accuracy of approximately 99.97%.It also recorded strong performance in other metrics,achieving 93%precision,100%sensitivity,89%specificity,and an AUC-ROC score of 97%.Comparative analysis with baseline feature engineering methods confirmed the superior efficacy of HCL Net.The model significantly reduces misclassification,particularly between honeycombing and GGO lungs,enhancing diagnostic precision and reliability in lung image analysis.
基金supported by the Special Fund for Scientific and Technological Innovation Strategy of Guangdong Province(2022A0505030025)the Science and Technology Fund,FDCT,Macao SAR(0064/2021/A2)
文摘This paper addresses the verification of strong currentstate opacity with respect to real-time observations generated from a discrete-event system that is modeled with time labeled Petri nets. The standard current-state opacity cannot completely characterize higher-level security. To ensure the higher-level security requirements of a time-dependent system, we propose a strong version of opacity known as strong current-state opacity. For any path(state-event sequence with time information)π derived from a real-time observation that ends at a secret state, the strong current-state opacity of the real-time observation signifies that there is a non-secret path with the same real-time observation as π. We propose general and non-secret state class graphs, which characterize the general and non-secret states of time-dependent systems, respectively. To capture the observable behavior of non-secret states, a non-secret observer is proposed.Finally, we develop a structure called a real-time concurrent verifier to verify the strong current-state opacity of time labeled Petri nets. This approach is efficient since the real-time concurrent verifier can be constructed by solving a certain number of linear programming problems.
基金Supported by Noor Ophthalmology Research Center.
文摘This study aimed to differentiate between the therapeutic and purely cosmetic aspects of keratopigmentation by examining the results of previous studies.Corneal opacity,corneal leukoma,and iris abnormalities represent the primary reasons for keratopigmentation in therapeutic settings.In therapeutic contexts,the most commonly reported adverse effects included inflammation,foreign body sensation,light sensitivity,reduced intraocular pressure,and neovascularization.The side effects most frequently noted in purely cosmetic applications were photophobia and ectasia.In therapeutic scenarios,patients who experience pre-operative vision impairment often report not only the absence of visual loss post-procedure but also an enhancement in visual acuity,accompanied by alleviation of symptoms such as glare.In purely cosmetic procedures,visual acuity tends to remain stable.Research indicates that dissatisfaction with keratopigmentation results is more prevalent in studies aimed at therapeutic interventions compared to cosmetic enhancements.Furthermore,instances of discoloration and fading are more frequently observed in therapeutic cases than in cosmetic ones,including the need for retouching.The use of keratopigmentation in both therapeutic and cosmetic settings produces positive and acceptable results,especially in cosmetic applications.Nevertheless,complications are more commonly observed in therapeutic situations,which underscores the importance of candidate selection for keratopigmentation procedures.
基金supported by the Social Science Foundation of Liaoning Province(L23BJY022).
文摘We explored the effects of algorithmic opacity on employees’playing dumb and evasive hiding rather than rationalized hiding.We examined the mediating role of job insecurity and the moderating role of employee-AI collaboration.Participants were 421 full-time employees(female=46.32%,junior employees=31.83%)from a variety of organizations and industries that interact with AI.Employees filled out data on algorithm opacity,job insecurity,knowledge hiding,employee-AI collaboration,and control variables.The results of the structural equation modeling indicated that algorithm opacity exacerbated employees’job insecurity,and job insecurity mediated between algorithm opacity and playing dumb and evasive hiding rather than rationalized hiding.The relationship between algorithmic opacity and playing dumb and evasive hiding was more positive when the level of employee-AI collaboration was higher.These findings suggest that employee-AI collaboration reinforces the indirect relationship between algorithmic opacity and playing dumb and evasive hiding.Our study contributes to research on human and AI collaboration by exploring the dark side of employee-AI collaboration.
文摘Objective: To analyze the clinical manifestations of intraocular lens (IOL) opacity after cataract surgery through case reports, and to explore its pathogenesis and diagnosis and treatment ideas, so as to provide a basis for the early diagnosis and correct treatment of IOL opacity. Methods: The clinical data of one patient diagnosed with IOL opacity and underwent intraocular lens replacement in the Department of Ophthalmology, Affiliated Hospital of Youjiang Medical College for Nationalities in December 2023 were reported. The characteristics of IOL opacity were observed, and the research progress and pathogenesis of IOL opacity were understood by consulting the literature. Results: This patient is the first case of IOL opacity in our hospital. The specific reason is unclear. It is considered to be related to the IOL material. Conclusion: Hydrophilic acrylic IOL is widely used in clinic because of its good histocompatibility. However, due to its hydrophilicity, there are more cases of IOL opacity than other types of IOL. At present, there is no unified conclusion on the etiology and mechanism of IOL opacity. IOL opacity can seriously affect vision and is easily misdiagnosed as a posterior cataract. We should fully disperse the large pupil and carefully observe under the slit lamp. The most effective treatment for this disease is IOL replacement.
