Objective Against the backdrop of rapid advancements in artificial intelligence(AI)technology,generative AI such as DeepSeek and Manus is integrating into and reshaping the healthcare industry with disruptive force.AI and related technologies are increasingly being applied in hospital management.Based on the mature application of Robotic Process Automa-tion(RPA)technology,this study aims to explore the potential of AI technology in the automatic collection,processing,and dis-tribution of daily hospital information statistics reports.By constructing an automated daily report generation system,it seeks to address issues in traditional hospital management such as high labor costs and low efficiency,thereby providing technical support for the construction of smart hospitals.Methods This study took a tertiary hospital in Shenzhen as an empirical research sub-ject,developing a specific workflow plan for automatically generating daily hospital information statistics reports based on AI+RPA technology to empower smart hospital management.To tackle data missing issues,an innovative 30-minute silent period manual review mechanism was established,creating a dual-guarantee system of"automatic machine processing+manual seconda-ry verification."The study focused on achieving technical breakthroughs in key links including data acquisition,storage and pro-cessing,daily report generation,and scheduled push notifications,ensuring the automation and accuracy of the daily hospital in-formation statistics reporting process.Results The AI+RPA technology demonstrated excellent implementation outcomes in key aspects and specific application scenarios such as data acquisition,storage,automated daily report generation,and automated sending to designated locations for hospital information statistics.The efficiency of daily report generation was significantly im-proved,labor costs were reduced,and data accuracy was enhanced.The human-machine collaboration mechanism effectively re-solved low-level errors caused by missing data and optimized the overall daily report generation process.Conclusion By recon-structing the workflow for daily hospital information statistics reports,AI+RPA technology significantly enhanced management ef-ficiency and data quality,meeting the data needs of hospital management departments.Compared to the daily consumption of hu-man resources previously,the"Statistics Robot"not only substantially reduces labor costs but can also be scheduled to run auto-matically during idle periods for both personnel and information systems.It breaks down departmental barriers and data silos,a-chieving higher resource allocation efficiency and laying a data foundation for smart hospital development.In the future,the deep integration of AI with RPA technology and the targeted development of software robots for other specialized purposes will bring more innovation and transformation to smart hospitals,further enhancing hospital management efficiency and service quality.

Objective Against the backdrop of rapid advancements in artificial intelligence(AI)technology,generative AI such as DeepSeek and Manus is integrating into and reshaping the healthcare industry with disruptive force.AI and related technologies are increasingly being applied in hospital management.Based on the mature application of Robotic Process Automa-tion(RPA)technology,this study aims to explore the potential of AI technology in the automatic collection,processing,and dis-tribution of daily hospital information statistics reports.By constructing an automated daily report generation system,it seeks to address issues in traditional hospital management such as high labor costs and low efficiency,thereby providing technical support for the construction of smart hospitals.Methods This study took a tertiary hospital in Shenzhen as an empirical research sub-ject,developing a specific workflow plan for automatically generating daily hospital information statistics reports based on AI+RPA technology to empower smart hospital management.To tackle data missing issues,an innovative 30-minute silent period manual review mechanism was established,creating a dual-guarantee system of"automatic machine processing+manual seconda-ry verification."The study focused on achieving technical breakthroughs in key links including data acquisition,storage and pro-cessing,daily report generation,and scheduled push notifications,ensuring the automation and accuracy of the daily hospital in-formation statistics reporting process.Results The AI+RPA technology demonstrated excellent implementation outcomes in key aspects and specific application scenarios such as data acquisition,storage,automated daily report generation,and automated sending to designated locations for hospital information statistics.The efficiency of daily report generation was significantly im-proved,labor costs were reduced,and data accuracy was enhanced.The human-machine collaboration mechanism effectively re-solved low-level errors caused by missing data and optimized the overall daily report generation process.Conclusion By recon-structing the workflow for daily hospital information statistics reports,AI+RPA technology significantly enhanced management ef-ficiency and data quality,meeting the data needs of hospital management departments.Compared to the daily consumption of hu-man resources previously,the"Statistics Robot"not only substantially reduces labor costs but can also be scheduled to run auto-matically during idle periods for both personnel and information systems.It breaks down departmental barriers and data silos,a-chieving higher resource allocation efficiency and laying a data foundation for smart hospital development.In the future,the deep integration of AI with RPA technology and the targeted development of software robots for other specialized purposes will bring more innovation and transformation to smart hospitals,further enhancing hospital management efficiency and service quality.

Objective To investigate the clinical characteristics of chronic invasive fungal rhinosinusitis.Methods Clinical features and outcomes of 46 proven cases of chronic invasive fungal rhinosinusitis admitted in Huashan Hospital,Fudan University from January 2009 to December 2016 were retrospectively reviewed.Results Of the 46 patients enrolled,left sphenoid sinus,ethmoid sinus and maxillary sinus were affected in 24,23 and 20 cases,respectively,while right maxillary sinus,ethmoid sinus and sphenoid sinus were affected in 18,16 and 15 eases,respectively.Left and right frontal sinus were affected in 9 and 6 cases,respectively.The central nervous system and orbit were the most commonly affected sites in external nasal involvements,noted in 22 cases respectively.Left sphenoid (17 cases) and ethmoid sinus (15 cases) involvements were most common in central nervous system affected patients.Left sphenoid (14 cases) and ethmoid sinus (13 cases) involvements were most common in orbit affected patients.Aspergillus species were the primary pathogens observed in 42 eases.Zygomycete,candida and dark filamentous fungus were observed in two,one and one case,respectively.Pathologically,37 of the cases were chronic nongranulomatous type and the left 9 were chronic granulomatous type who were all immunocompetent hosts.The initial symptoms usually included headache,dizziness and nasal discomforts including nasal obstruction and purulent secretion.The chief complaints usually included headache,dizziness,and visual disturbances including blurred vision,vision loss or even blindness.Antifungal treatment combined with surgical interventions for removal or drainage focus lesions achieved significant effect,and 42 patients were cured.Conclusions Chronic invasive fungal rhinosinusitis should be taken into consideration in the presence of nasal discomforts or nonspecific symptoms such as headache and dizziness.The possibility of chronic invasive fungal rhinosinusitis should be cautious after the emergence of vision abnormalities.

Objective To evaluate the relationship between CD4+ T lymphocyte count and results of enzyme-linked immunospot (ELISPOT) assay in human immunodeficiency virus (HIV)-Mycobacterium tuberculosis (M.tb) coinfected patients.Methods A total of 193 HIV-infected individuals in Yunnan Province and Shanghai were enrolled.T-SPOT.TB assay was employed to detect M.tb specific T lymphocyte in the peripheral blood mononuclear cells (PBMC).CD4+ T lymphocyte in PBMC from the enrolled subjects was detected by flow cytometry.Data were analyzed using t test.ResultsThe incidence of latent tuberculosis in HIV-infected individuals was 30.6％.The CD4+ T lymphocyte counts in HIV-infected individuals with active tuberculosis were 190×106/L,which were significantly lower than those in HIV-infected individuals with latent tuberculosis (484×106/L; t=6.665,P＜0.01).The HIV-infected individuals were stratified according to CD4+ T lymphocyte counts of ＞500×106/L,200×106-500×106/L,and ＜200×106/L and the constituent ratios of active tuberculosis/latent tuberculosis were 1∶16.2,1∶1.3 and 5.6∶1,respectively.Among 79 subjects with positive T-SPOT.TB results,20 were coinfected with active tuberculosis,in which 14 had CD4+ T lymphocyte counts of ＜200 ×106/L,5 had 200×105-500×106/L and 1 had ＞500×106/L.Fifty-two in 59 HIV/latent tuberculosis patients individuals had CD4+ T lymphocyte counts of ＞200×106/L.ConclusionsThe prevalence of latent tuberculosis in HIV-infected individuals is high in China.Cellular immunity in HIV-infected individuals with active tuberculosis is severely impaired.With the decrease of CD4 ′ T lymphocyte counts,patients with latent tuberculosis are prone to develop active tuberculosis in HIV-infected individuals.The negative predictive value of T-SPOT.TB is significantly diminished in patient with low CD4+ T lymphocyte counts,especially less than 200×106/L.

Objective To study the clinical features and antifungal therapeutic effects in nonacquired immune deficiency syndrome(AIDS)patients with cryptococcal meningitis. Methods One hundred and fifty-four non-AIDS patients with cryptococcal meningitis admitted to Huashan Hospital, Fudan University from 1997 to 2007 were reviewed retrospectively. Clinical characteristics, initial antifungal therapies and outcome of these patients were analyzed. Continuous variables were analyzed using t test and categorical variables were compared by X~2 test or Fisher's exact test. Kaplan-Meier survival curves of different therapies were compared with log-rank test. Results Fifty-one patients (33.12%)had one or more predisposing factors. Headache, fever, meningeal irritation, vomiting and altered mental status were common clinical symptoms and signs during the course of diseases. The positive rates of cerebrospinal fluid(CSF)smear, CSF culture and detection of CSF cryptococcal capsular polysaccharide antigen were 88.44%,78.95%and 100.00%,respectively.Twelve cases were excluded because treatment durations were less than 7 days, including 9 died,2 discharged against medical advice due to illness exacerbation and 1 lost after against medical advice discharge. The remaining 142 patients were evaluated for therapeutic effects. The effective rates in amphotericin B (AmB)group, fluconazole group and AmB plus fluconazole group were 78.3%(36/46),33.3%(8/24)and 76.0%(38/50),respectively. The therapeutic effects in AmB group and AmB plus fluconazole group were superior to fluconazole group(X~2=13.6354,12.5509;P＜0.01).Eleven patients were lost during 1-year follow-up. The attributable and overall mortality in the remaining 143 patients were 19.58% and 28.67%,respectively.The 1-year survival rates in AmB group and AmB plus fluconazole group were significantly higher than that in fluconazole group. Conclusions The mortality of non-AIDS cryptococcal meningitis is still high,which is closely correlated with initial antifungal therapies. AmB alone or combined with flucytosine is related to both higher successful response and higher survival rate, while the efficacy of initial fluconazole alone or combined with flucytosine is poor.