Medical cross-modal retrieval aims to achieve semantic similarity search between different modalities of medical cases, such as quickly locating relevant ultrasound images through ultrasound reports, or using ultrasound images to retrieve matching reports. However, existing medical cross-modal hash retrieval methods face significant challenges, including semantic and visual differences between modalities and the scalability issues of hash algorithms in handling large-scale data. To address these challenges, this paper proposes a Medical image Semantic Alignment Cross-modal Hashing based on Transformer (MSACH). The algorithm employed a segmented training strategy, combining modality feature extraction and hash function learning, effectively extracting low-dimensional features containing important semantic information. A Transformer encoder was used for cross-modal semantic learning. By introducing manifold similarity constraints, balance constraints, and a linear classification network constraint, the algorithm enhanced the discriminability of the hash codes. Experimental results demonstrated that the MSACH algorithm improved the mean average precision (MAP) by 11.8% and 12.8% on two datasets compared to traditional methods. The algorithm exhibits outstanding performance in enhancing retrieval accuracy and handling large-scale medical data, showing promising potential for practical applications.
Algorithms ; Semantics ; Humans ; Ultrasonography ; Information Storage and Retrieval/methods* ; Image Processing, Computer-Assisted/methods*

Online hashing methods are receiving increasing attention in cross modal medical image retrieval research. However, existing online methods often lack the learning ability to maintain semantic correlation between new and existing data. To this end, we proposed online semantic similarity cross-modal hashing (OSCMH) learning framework to incrementally learn compact binary hash codes of medical stream data. Within it, a sparse representation of existing data based on online anchor datasets was designed to avoid semantic forgetting of the data and adaptively update hash codes, which effectively maintained semantic correlation between existing and arriving data and reduced information loss as well as improved training efficiency. Besides, an online discrete optimization method was proposed to solve the binary optimization problem of hash code by incrementally updating hash function and optimizing hash code on medical stream data. Compared with existing online or offline hashing methods, the proposed algorithm achieved average retrieval accuracy improvements of 12.5% and 14.3% on two datasets, respectively, effectively enhancing the retrieval efficiency in the field of medical images.
Semantics ; Humans ; Algorithms ; Information Storage and Retrieval/methods* ; Diagnostic Imaging ; Image Processing, Computer-Assisted/methods*

Data space, as an innovative data management and sharing model, is emerging in the medical and health sectors. This study expounds on the conceptual connotation of data space and delineates its key technologies, including distributed data storage, standardization and interoperability of data sharing, data security and privacy protection, data analysis and mining, and data space assessment. By analyzing the real-world cases of data spaces within medicine and health, this study compares the similarities and differences across various dimensions such as purpose, architecture, data interoperability, and privacy protection. Meanwhile, data spaces in these fields are challenged by the limited computing resources, the complexities of data integration, and the need for optimized algorithms. Additionally, legal and ethical issues such as unclear data ownership, undefined usage rights, risks associated with privacy protection need to be addressed. The study notes organizational and management difficulties, calling for enhancements in governance framework, data sharing mechanisms, and value assessment systems. In the future, technological innovation, sound regulations, and optimized management will help the development of the medical and health data space. These developments will enable the secure and efficient utilization of data, propelling the medical industry into an era characterized by precision, intelligence, and personalization.
Humans ; Computer Security ; Information Dissemination ; Data Management ; Information Storage and Retrieval ; Data Mining

INTRODUCTION: Hospitalisations can pose hazards and may not be an appropriate care setting for frail nursing home (NH) residents. Few studies have quantified the extent of NH resident hospitalisations in Singapore, hence we aimed to address this knowledge gap by studying characteristics of unplanned hospitalisations over a 1-year period. METHOD: This was a retrospective cohort study of 9922 subsidised residents across 59 NHs in Singapore, with analysis using administrative healthcare data. Key measures included inpatient admission and emergency department visit rates, final discharge diagnoses and estimated costs. We examined correlates of inpatient admissions with a multivariable zero-inflated negative binomial regression model incorporating demogra-phics, institutional characteristics and Charlson Comorbidity Index. RESULTS: There were 6620 inpatient admissions in 2015, equivalent to 2.23 admissions per 1000 resident days, and the majority were repeat admissions (4504 admissions or 68.0%). Male sex (incidence rate ratio [IRR] 1.23), approaching end-of-life (IRR 2.14), hospitalisations in the past year (IRR 2.73) and recent NH admission within the last 6 months (IRR 1.31-1.99) were significantly associated with inpatient admission rate. Top 5 discharge diagnoses were lower respiratory tract infections (27.3%), urinary tract infection (9.3%), sepsis (3.1%), cellulitis (1.9%) and gastroenteritis (1.1%). We estimated the total system cost of admissions of subsidised residents to be SGD40.2 million (USD29.1 million) in 2015. CONCLUSION We anticipate that unplanned hospitali-sation rate will increase over time, especially with an increasing number of residents who will be cared for in NHs. Our findings provide a baseline to inform stakeholders and develop strategies to address this growing problem.
Humans ; Singapore ; Nursing Homes/economics* ; Retrospective Studies ; Male ; Female ; Hospitalization/statistics & numerical data* ; Aged ; Aged, 80 and over ; Emergency Service, Hospital/statistics & numerical data* ; Information Storage and Retrieval ; Homes for the Aged/economics* ; Frail Elderly/statistics & numerical data*

Nowadays, the massive medical data have already influenced the information construction in medical institutes, so it is not enough to solely rely on traditional local storage system to solve the problems like the read/write speed, visualization, and economy brought about by the massive data. Furthermore, various medical cloud services have been developed at home and abroad that patients' medical data can be shared through all medical institutes on the cloud, which makes a higher demand on the transmission speed of the medical data. This article analyzes from multiple aspects like high availability and costs by performing a medical image transmission speed test on the three mainstream storage technologies to provide an optional storage system for future medical image big data in the access process. The experimental result shows that it can be found that in the process of accessing medical image big data, the access speed and performance of object storage system is better than those of the existing local storage systems. However, with comprehensive consideration, it is recommended that the distributed file storage system like HDFS be the first choice for the storage system of the medical images.
Cloud Computing ; Computer Communication Networks ; Humans ; Information Storage and Retrieval ; Technology

Post-marketing surveillance is the principal means to ensure drug use safety. The spontaneous report is the essential method of post-marketing surveillance for drug safety. Often, most spontaneous reports come from medical staff and sometimes come from patients who use the drug. The posts published by individuals on social media platforms that contain drugs and related adverse reaction content have gradually been seen as a new data source similar to spontaneous reports from drug users in recent years. Those user-generated posts potentially provide researchers and regulators with new opportunities to conduct post-marketing surveillance for drug safety from patients' perspectives mostly rather than medical professionals and can afford the possibility theoretically to discover drug-related safety issues earlier than traditional methods. Social media data as a new data source for safety signal detection and signal reinforcement have the unique advantages, such as population coverage, type of drugs, type of adverse reactions, data timeliness and quantity. Most of the social media data used in post-marketing surveillance research for drug safety are still text data in English, and even multiple languages are used by different people worldwide on several social media platforms. Unfortunately, there is still a controversy in the academic circles whether social media data can be used as reliable data sources for routine post-marketing surveillance for drug safety. A couple of obstacles of data, methods and ethics must be overcome before leveraging social media data for post-marketing surveillance. The number of Chinese social media users is large, and the social media data in the Chinese language is rapidly snowballing, which can be employed as the potential data source for post-marketing surveillance for drug safety. However, due to the Chinese language's specific characteristics, the text's diversity is different from the English text, and there is not enough accepted corpus in medical scenarios. Besides, the lack of domestic laws and regulations on privacy and security protection of social media data poses more challenges for applying Chinese social media data for post-market surveillance. The significance of social media data to post-marketing surveillance for drug safety is undoubtedly significant. It will be an essential development direction for future research to overcome the challenges of using social media data by developing new technologies and establishing new mechanisms.
Humans ; Information Storage and Retrieval ; Marketing ; Social Media

The clinical comprehensive evaluation of drugs is an important basis for the return of clinical value, decision-making of medical and health authorities, and allocation of medical resources. In July 2021, the National Health Commission issued the Guidelines for the Management of Clinical Comprehensive Evaluation of Drugs(trial version 2021), which required the evaluation to be implemented from the six dimensions(safety, effectiveness, economy, innovation, suitability, and accessibility), and made detailed arrangements for the clinical comprehensive evaluation of drugs. As Chinese patent medicine differs from chemical medicines in terms of effective components and action modes, the clinical comprehensive evaluation of Chinese patent medicine should highlight the characteristics and advantages of traditional Chinese medicine(TCM) on the basis of general requirements of comprehensive clinical evaluation of drugs. At present, in the clinical comprehensive evaluation of Chinese patent medicine, unified report standards have not yet been generated, resulting in the uneven quality of existing reports. To standardize the clinical comprehensive evaluation report of Chinese patent medicine and improve its quality, the editorial team, based on the relevant policy documents of clinical comprehensive evaluation of drugs, formulated the clinical comprehensive evaluation report standards for Chinese patent medicine in combination with the previous practice and expert opinions. The report standards, containing seven sections with 15 items determined, focus on data source, evaluation content, evidence synthesis, quality control, and evaluation results supported with detailed interpretations to help researchers better understand and apply the report standards for clinical comprehensive evaluation of Chinese patent medicine, improve the report quality, and provide references for the decision-making by the national medical management authorities.
China ; Drugs, Chinese Herbal ; Information Storage and Retrieval ; Medicine, Chinese Traditional ; Nonprescription Drugs ; Quality Control

This article is based on basic data such as field surveys and literature surveys, contrasting and analyzing the distribution of Callicarpa nudiflora by different zoning methods, different data sources, and different spatial scales. The results showed that there were certain differences in the distribution results obtained by using different methods, such as qualitative description, similar ecological environment, and niche model, to divide the distribution of the C. nudiflora, but all of them could reflect the distribution of C. nudiflora to different degrees. Among them, the qualitative description division method has certain advantages in macro guidance in a large scale. The distribution range obtained by the ecological environment similar division method is wider than that obtained by applying the qualitative description method and the niche model method. The results of the zoning of the distribution of the C. nudiflora obtained from different data sources were different. The number and representativeness of the survey data have an impact on the zoning results. Through the analysis of the distribution of different spatial scales, the ecological factors and contribution rates that affect the distribution of C. nudiflora are different in China and in the world. The comprehensive multi-source data analysis showed that C. nudiflora mainly distributed in southern coastal provinces such as Hainan, Guangdong, Guangxi and Fujian in China, and also in Jiangxi, Guizhou, Yunnan, Sichuan, Chongqing, Hunan, Gansu, Taiwan and other provinces. Globally, C. nudiflora are suitable for distribution in Southeast Asia, such as China, Vietnam, Laos, Myanmar, India, etc. There are also potential distribution areas in the southern United States and Mexico.
Callicarpa ; China ; Data Collection ; Information Storage and Retrieval ; Vietnam

The burgeoning wealth of available scientific information – aided in part by (a) expansion in the definition of “literature”, (b) dramatic increase in the scientific output available for the scientific community's perusal, and (c) ease of access afforded by various databases and search engines – poses several challenges to researchers and to the credibility of their research findings. One method to discourage reference to “fraudulent, incomplete, or obsolete data” in the literature is citations searching. This paper presented a short overview of citations searching, its advantages and disadvantages, as well as its implications for stakeholders in the academic community.
Information Storage and Retrieval ; Research Design ; Review Literature as Topic

OBJECTIVES: Early initiation of breastfeeding (EIBF) is one of the most cost-effective strategies to reduce neonatal mortality. We sought to determine the prevalence and determinants of EIBF in Peru.METHODS: We performed a cross-sectional analytical study of the 2018 Peruvian Demographic and Family Health Survey as a secondary data source. In total, 19,595 children born during the 5 years prior to the survey were included in the study. The dependent variable (EIBF status), socio-demographic variables, and pregnancy-related variables were analyzed using a multivariate logistic regression model to identify the determinants of EIBF.RESULTS: The prevalence of EIBF in the study population was 49.7%. Cesarean deliveries were associated with a lower likelihood of EIBF (adjusted odds ratio [aOR], 0.06; 95% confidence interval [CI], 0.05 to 0.07) than were vaginal deliveries. Newborns born at public health centers (aOR, 1.37; 95% CI, 1.15 to 1.65) had a higher rate of EIBF than those not born at public or private health centers. Women from the jungle region (aOR, 2.51; 95% CI, 2.17 to 2.89) had higher odds of providing EIBF than those from the coast. Mothers with more than a secondary education (aOR, 0.65; 95% CI, 0.55 to 0.76) were less likely to breastfeed during the first hour of the newborn's life than women with primary or no education.CONCLUSIONS: More than half of Peruvian children do not breastfeed during the first hour after birth. The major determinants of EIBF status were the delivery mode and the region of maternal residence. Strategies are needed to promote early breastfeeding practices.
Breast Feeding ; Child ; Education ; Family Health ; Female ; Health Surveys ; Humans ; Infant ; Infant Mortality ; Infant, Newborn ; Information Storage and Retrieval ; Logistic Models ; Mothers ; Odds Ratio ; Parturition ; Peru ; Prevalence ; Public Health