Diagnosis and detection of pneumonia using weak-label based on X-ray images: a multi-center study.

PURPOSE: Development and assessment the deep learning weakly supervised algorithm for the classification and detection pneumonia via X-ray. METHODS: This retrospective study analyzed two publicly available dataset that contain X-ray images of pneumonia cases and normal cases. The first dataset from Guangzhou Women and Children's Medical Center. It contains a total of 5,856 X-ray images, which are divided into training, validation, and test sets with 8:1:1 ratio for algorithm training and testing. The deep learning algorithm ResNet34 was employed to build diagnostic model. And the second public dataset were collated by researchers from Qatar University and the University of Dhaka along with collaborators from Pakistan and Malaysia and some medical doctors. A total of 1,300 images of COVID-19 positive cases, 1,300 normal images and 1,300 images of viral pneumonia for external validation. Class activation map (CAM) were used to location the pneumonia lesions. RESULTS: The ResNet34 model for pneumonia detection achieved an AUC of 0.9949 [0.9910-0.9981] (with an accuracy of 98.29% a sensitivity of 99.29% and a specificity of 95.57%) in the test dataset. And for external validation dataset, the model obtained an AUC of 0.9835[0.9806-0.9864] (with an accuracy of 94.62%, a sensitivity of 92.35% and a specificity of 99.15%). Moreover, the CAM can accurately locate the pneumonia area. CONCLUSION: The deep learning algorithm can accurately detect pneumonia and locate the pneumonia area based on weak supervision information, which can provide potential value for helping radiologists to improve their accuracy of detection pneumonia patients through X-ray images.

[Design and implementation of Internet of Things for emergency medical devices based on cloud-edge-device architecture].

Internet of Things (IoT) technology plays an important role in smart healthcare. This paper discusses IoT solution for emergency medical devices in hospitals. Based on the cloud-edge-device architecture, different medical devices were connected; Streaming data were parsed, distributed, and computed at the edge nodes; Data were stored, analyzed and visualized in the cloud nodes. The IoT system has been working steadily for nearly 20 months since it run in the emergency department in January 2021. Through preliminary analysis with collected data, IoT performance testing and development of early warning model, the feasibility and reliability of the in-hospital emergency medical devices IoT was verified, which can collect data for a long time on a large scale and support the development and deployment of machine learning models. The paper ends with an outlook on medical device data exchange and wireless transmission in the IoT of emergency medical devices, the connection of emergency equipment inside and outside the hospital, and the next step of analyzing IoT data to develop emergency intelligent IoT applications.

[Development of intelligent monitoring system based on Internet of Things and wearable technology and exploration of its clinical application mode].

Wearable monitoring, which has the advantages of continuous monitoring for a long time with low physiological and psychological load, represents a future development direction of monitoring technology. Based on wearable physiological monitoring technology, combined with Internet of Things (IoT) and artificial intelligence technology, this paper has developed an intelligent monitoring system, including wearable hardware, ward Internet of Things platform, continuous physiological data analysis algorithm and software. We explored the clinical value of continuous physiological data using this system through a lot of clinical practices. And four value points were given, namely, real-time monitoring, disease assessment, prediction and early warning, and rehabilitation training. Depending on the real clinical environment, we explored the mode of applying wearable technology in general ward monitoring, cardiopulmonary rehabilitation, and integrated monitoring inside and outside the hospital. The research results show that this monitoring system can be effectively used for monitoring of patients in hospital, evaluation and training of patients' cardiopulmonary function, and management of patients outside hospital.

[Quantitative analysis of breathing patterns based on wearable systems].

Breathing pattern parameters refer to the characteristic pattern parameters of respiratory movements, including the breathing amplitude and cycle, chest and abdomen contribution, coordination, etc. It is of great importance to analyze the breathing pattern parameters quantificationally when exploring the pathophysiological variations of breathing and providing instructions on pulmonary rehabilitation training. Our study provided detailed method to quantify breathing pattern parameters including respiratory rate, inspiratory time, expiratory time, inspiratory time proportion, tidal volume, chest respiratory contribution ratio, thoracoabdominal phase difference and peak inspiratory flow. We also brought in "respiratory signal quality index" to deal with the quality evaluation and quantification analysis of long-term thoracic-abdominal respiratory movement signal recorded, and proposed the way of analyzing the variance of breathing pattern parameters. On this basis, we collected chest and abdomen respiratory movement signals in 23 chronic obstructive pulmonary disease (COPD) patients and 22 normal pulmonary function subjects under spontaneous state in a 15 minute-interval using portable cardio-pulmonary monitoring system. We then quantified subjects' breathing pattern parameters and variability. The results showed great difference between the COPD patients and the controls in terms of respiratory rate, inspiratory time, expiratory time, thoracoabdominal phase difference and peak inspiratory flow. COPD patients also showed greater variance of breathing pattern parameters than the controls, and unsynchronized thoracic-abdominal movements were even observed among several patients. Therefore, the quantification and analyzing method of breathing pattern parameters based on the portable cardiopulmonary parameters monitoring system might assist the diagnosis and assessment of respiratory system diseases and hopefully provide new parameters and indexes for monitoring the physical status of patients with cardiopulmonary disease.

[Investigation on new paradigm of clinical physiological monitoring by using wearable devices].

As a low-load physiological monitoring technology, wearable devices can provide new methods for monitoring, evaluating and managing chronic diseases, which is a direction for the future development of monitoring technology. However, as a new type of monitoring technology, its clinical application mode and value are still unclear and need to be further explored. In this study, a central monitoring system based on wearable devices was built in the general ward (non-ICU ward) of PLA General Hospital, the value points of clinical application of wearable physiological monitoring technology were analyzed, and the system was combined with the treatment process and applied to clinical monitoring. The system is able to effectively collect data such as electrocardiogram, respiration, blood oxygen, pulse rate, and body position/movement to achieve real-time monitoring, prediction and early warning, and condition assessment. And since its operation from March 2018, 1 268 people (657 patients) have undergone wearable continuous physiological monitoring until January 2020, with data from a total of 1 198 people (632 cases) screened for signals through signal quality algorithms and manual interpretation were available for analysis, accounting for 94.48 % (96.19%) of the total. Through continuous physiological data analysis and manual correction, sleep apnea event, nocturnal hypoxemia, tachycardia, and ventricular premature beats were detected in 232 (36.65%), 58 (9.16%), 30 (4.74%), and 42 (6.64%) of the total patients, while the number of these abnormal events recorded in the archives was 4 (0.63%), 0 (0.00%), 24 (3.80%), and 15 (2.37%) cases. The statistical analysis of sleep apnea event outcomes revealed that patients with chronic diseases were more likely to have sleep apnea events than healthy individuals, and the incidence was higher in men (62.93%) than in women (37.07%). The results indicate that wearable physiological monitoring technology can provide a new monitoring mode for inpatients, capturing more abnormal events and provide richer information for clinical diagnosis and treatment through continuous physiological parameter analysis, and can be effectively integrated into existing medical processes. We will continue to explore the applicability of this new monitoring mode in different clinical scenarios to further enrich the clinical application of wearable technology and provide richer tools and methods for the monitoring, evaluation and management of chronic diseases.

Construction and Application of a Medical-Grade Wireless Monitoring System for Physiological Signals at General Wards.

Physiological signals can contain abundant personalized information and indicate health status and disease deterioration. However, in current medical practice, clinicians working in the general wards are usually lack of plentiful means and tools to continuously monitor the physiological signals of the inpatients. To address this problem, we here presented a medical-grade wireless monitoring system based on wearable and artificial intelligence technology. The system consists of a multi-sensor wearable device, database servers and user interfaces. It can monitor physiological signals such as electrocardiography and respiration and transmit data wirelessly. We highly integrated the system with the existing hospital information system and explored a set of processes of physiological signal acquisition, storage, analysis, and combination with electronic health records. Multi-scale information extracted from physiological signals and related to the deterioration or abnormality of patients could be shown on the user interfaces, while a variety of reports could be provided daily based on time-series signal processing technology and machine learning to make more information accessible to clinicians. Apart from an initial attempt to implement the system in a realistic clinical environment, we also conducted a preliminary validation of the core processes in the workflow. The heart rate veracity validation of 22 patient volunteers showed that the system had a great consistency with ECG Holter, and bias for heart rate was 0.04 (95% confidence interval: -7.34 to 7.42) beats per minute. The Bland-Altman analysis showed that 98.52% of the points were located between Mean ± 1.96SD. This system has been deployed in the general wards of the Hyperbaric Oxygen Department and Respiratory Medicine Department and has collected more than 1000 cases from the clinic. The whole system will continue to be updated based on clinical feedback. It has been demonstrated that this system can provide reliable physiological monitoring for patients in general wards and has the potential to generate more personalized pathophysiological information related to disease diagnosis and treatment from the continuously monitored physiological data.

[Study on the accuracy of cardiopulmonary physiological measurements by a wearable physiological monitoring system under different activity conditions].

This paper aims to study the accuracy of cardiopulmonary physiological parameters measurement under different exercise intensity in the accompanying (wearable) physiological parameter monitoring system. SensEcho, an accompanying physiological parameter monitoring system, and CORTEX METALYZER 3B, a cardiopulmonary function testing system, were used to simultaneously collect the cardiopulmonary physiological parameters of 28 healthy volunteers (17 males and 11 females) in various exercise states, such as standing, lying down and Bruce treadmill exercise. Bland-Altman analysis, correlation analysis and other methods, from the perspective of group and individual, were used to contrast and analyze the two types of equipment to measure parameters of heart rate and breathing rate. The results of group analysis showed that the heart rate and respiratory rate data box charts collected by the two devices were highly consistent. The heart rate difference was (-0.407 ± 3.380) times/min, and the respiratory rate difference was (-0.560 ± 7.047) times/min. The difference was very small. The Bland-Altman plot of the heart rate and respiratory rate in each experimental stage showed that the proportion of mean ± 2SD was 96.86% and 95.29%, respectively. The results of individual analysis showed that the correlation coefficients of the whole-process heart rate and respiratory rate data were all greater than 0.9. In conclusion, SensEcho, as an accompanying physiological parameter monitoring system, can accurately measure the human heart rate, respiration rate and other key cardiopulmonary physiological parameters under various sports conditions. It can maintain good stability under various sports conditions and meet the requirements of continuous physiological signal collection and analysis application under sports conditions.

[Design and preliminary validation of a ubiquitous and wearable physiological monitoring system].

To achieve continuously physiological monitoring on hospital inpatients, a ubiquitous and wearable physiological monitoring system SensEcho was developed. The whole system consists of three parts: a wearable physiological monitoring unit, a wireless network and communication unit and a central monitoring system. The wearable physiological monitoring unit is an elastic shirt with respiratory inductive plethysmography sensor and textile electrocardiogram (ECG) electrodes embedded in, to collect physiological signals of ECG, respiration and posture/activity continuously and ubiquitously. The wireless network and communication unit is based on WiFi networking technology to transmit data from each physiological monitoring unit to the central monitoring system. A protocol of multiple data re-transmission and data integrity verification was implemented to reduce packet dropouts during the wireless communication. The central monitoring system displays data collected by the wearable system from each inpatient and monitors the status of each patient. An architecture of data server and algorithm server was established, supporting further data mining and analysis for big medical data. The performance of the whole system was validated. Three kinds of tests were conducted: validation of physiological monitoring algorithms, reliability of the monitoring system on volunteers, and reliability of data transmission. The results show that the whole system can achieve good performance in both physiological monitoring and wireless data transmission. The application of this system in clinical settings has the potential to establish a new model for individualized hospital inpatients monitoring, and provide more precision medicine to the patients with information derived from the continuously collected physiological parameters.

[Construction of multi-parameter emergency database and preliminary application research].

The analysis of big data in medical field cannot be isolated from the high quality clinical database, and the construction of first aid database in our country is still in the early stage of exploration. This paper introduces the idea and key technology of the construction of multi-parameter first aid database. By combining emergency business flow with information flow, an emergency data integration model was designed with reference to the architecture of the Medical Information Mart for Intensive Care III (MIMIC-III), created by Computational Physiology Laboratory of Massachusetts Institute of Technology (MIT), and a high-quality first-aid database was built. The database currently covers 22 941 medical records for 19 814 different patients from May 2015 to October 2017, including relatively complete information on physiology, biochemistry, treatment, examination, nursing, etc. And based on the database, the first First-Aid Big Data Datathon event, which 13 teams from all over the country participated in, was launched. The First-Aid database provides a reference for the construction and application of clinical database in China. And it could provide powerful data support for scientific research, clinical decision making and the improvement of medical quality, which will further promote secondary analysis of clinical data in our country.

[Research and Design of Respiratory Impedance Measurement System Based on Forced Oscillation Technique].

Forced oscillation technique(FOT)is an active method to test pulmonary function,which can derive the mechanical characteristics of the respiratory system with liner system identification theory by pushing in an oscillation air signal and measuring the changes of output pressure and flow.A pulmonary function determination system was developed based on the FOT in this paper.Several critical technologies of this determination system were analyzed,including the selection criteria of oscillation air generator,pressure and flow sensor,the signal design of oscillation air generator,and the synchronous sampling of pressure and flow data.A software program on LabVIEW platform was set up to control the determination system and get the measuring data.The performance of sensors and oscillation air generator was verified.According to the frequency response curve of the pressure,the amplitude of driving signal to the oscillation air generator was corrected at the frequency range between 4~40 Hz.A simulation experiment was carried out to measure the respiratory impedance of the active model lung ASL5000 and the results were close to the setting values of the model lung.The experiment testified that the pulmonary function determination system based on FOT had performance good enough to provide a tool for the in-depth research of the mechanical properties of the respiratory system.

[Evaluation of patient-ventilator synchrony of three new types of ventilators with pressure sunnort ventilation mode].

Pressure-support ventilation (PSV) is a form of important ventilation mode. Patient-ventilator synchrony of pressure support ventilation can be divided into inspiration-triggered and expiration-triggered ones. Whether the ventilator can track the patient's inspiration and expiration very well or not is an important evaluating item of the performance of the ventilator. The ventilator should response to the patient's inspiration effort on time and deliver the air flow to the patient under various conditions, such as different patient's lung types and inspiration effort, etc. Similarly, the ventilator should be able to response to the patient's expiration action, and to decrease the patient lung's internal pressure rapidly. Using the Active Servo Lung (ASL5000) respiratory simulation system, we evaluated the spontaneous breathing of PSV mode on E5, Servo i and Evital XL. The following parameters, the delay time before flow to the patient starts once the trigger variable signaling the start of inspiration, the lowest inspiratory airway pressure generated prior to the initiation of PSV, etc. were measured.

Illness severity assessment of older adults in critical illness using machine learning (ELDER-ICU): an international multicentre study with subgroup bias evaluation.

BACKGROUND: Comorbidity, frailty, and decreased cognitive function lead to a higher risk of death in elderly patients (more than 65 years of age) during acute medical events. Early and accurate illness severity assessment can support appropriate decision making for clinicians caring for these patients. We aimed to develop ELDER-ICU, a machine learning model to assess the illness severity of older adults admitted to the intensive care unit (ICU) with cohort-specific calibration and evaluation for potential model bias. METHODS: In this retrospective, international multicentre study, the ELDER-ICU model was developed using data from 14 US hospitals, and validated in 171 hospitals from the USA and Netherlands. Data were extracted from the Medical Information Mart for Intensive Care database, electronic ICU Collaborative Research Database, and Amsterdam University Medical Centers Database. We used six categories of data as predictors, including demographics and comorbidities, physical frailty, laboratory tests, vital signs, treatments, and urine output. Patient data from the first day of ICU stay were used to predict in-hospital mortality. We used the eXtreme Gradient Boosting algorithm (XGBoost) to develop models and the SHapley Additive exPlanations method to explain model prediction. The trained model was calibrated before internal, external, and temporal validation. The final XGBoost model was compared against three other machine learning algorithms and five clinical scores. We performed subgroup analysis based on age, sex, and race. We assessed the discrimination and calibration of models using the area under receiver operating characteristic (AUROC) and standardised mortality ratio (SMR) with 95% CIs. FINDINGS: Using the development dataset (n=50 366) and predictive model building process, the XGBoost algorithm performed the best in all types of validations compared with other machine learning algorithms and clinical scores (internal validation with 5037 patients from 14 US hospitals, AUROC=0·866 [95% CI 0·851-0·880]; external validation in the US population with 20 541 patients from 169 hospitals, AUROC=0·838 [0·829-0·847]; external validation in European population with 2411 patients from one hospital, AUROC=0·833 [0·812-0·853]; temporal validation with 4311 patients from one hospital, AUROC=0·884 [0·869-0·897]). In the external validation set (US population), the median AUROCs of bias evaluations covering eight subgroups were above 0·81, and the overall SMR was 0·99 (0·96-1·03). The top ten risk predictors were the minimum Glasgow Coma Scale score, total urine output, average respiratory rate, mechanical ventilation use, best state of activity, Charlson Comorbidity Index score, geriatric nutritional risk index, code status, age, and maximum blood urea nitrogen. A simplified model containing only the top 20 features (ELDER-ICU-20) had similar predictive performance to the full model. INTERPRETATION: The ELDER-ICU model reliably predicts the risk of in-hospital mortality using routinely collected clinical features. The predictions could inform clinicians about patients who are at elevated risk of deterioration. Prospective validation of this model in clinical practice and a process for continuous performance monitoring and model recalibration are needed. FUNDING: National Institutes of Health, National Natural Science Foundation of China, National Special Health Science Program, Health Science and Technology Plan of Zhejiang Province, Fundamental Research Funds for the Central Universities, Drug Clinical Evaluate Research of Chinese Pharmaceutical Association, and National Key R&D Program of China.

Clinically Interpretable Machine Learning Models for Early Prediction of Mortality in Older Patients with Multiple Organ Dysfunction Syndrome: An International Multicenter Retrospective Study.

BACKGROUND: Multiple organ dysfunction syndrome (MODS) is associated with a high risk of mortality among older patients. Current severity scores are limited in their ability to assist clinicians with triage and management decisions. We aim to develop mortality prediction models for older patients with MODS admitted to the ICU. METHODS: The study analyzed older patients from 197 hospitals in the United States and 1 hospital in the Netherlands. The cohort was divided into the young-old (65-80 years) and old-old (≥80 years), which were separately used to develop and evaluate models including internal, external, and temporal validation. Demographic characteristics, comorbidities, vital signs, laboratory measurements, and treatments were used as predictors. We used the XGBoost algorithm to train models, and the SHapley Additive exPlanations (SHAP) method to interpret predictions. RESULTS: Thirty-four thousand four hundred and ninety-seven young-old (11.3% mortality) and 21 330 old-old (15.7% mortality) patients were analyzed. Discrimination AUROC of internal validation models in 9 046 U.S. patients was as follows: 0.87 and 0.82, respectively; discrimination of external validation models in 1 905 EUR patients was as follows: 0.86 and 0.85, respectively; and discrimination of temporal validation models in 8 690 U.S. patients: 0.85 and 0.78, respectively. These models outperformed standard clinical scores like Sequential Organ Failure Assessment and Acute Physiology Score III. The Glasgow Coma Scale, Charlson Comorbidity Index, and Code Status emerged as top predictors of mortality. CONCLUSIONS: Our models integrate data spanning physiologic and geriatric-relevant variables that outperform existing scores used in older adults with MODS, which represents a proof of concept of how machine learning can streamline data analysis for busy ICU clinicians to potentially optimize prognostication and decision making.

A Simple Nomogram for Predicting Hospital Mortality of Patients Over 80 Years in ICU: An International Multicenter Retrospective Study.

OBJECTIVES: This study aimed to develop and validate an easy-to-use intensive care unit (ICU) illness scoring system to evaluate the in-hospital mortality for very old patients (VOPs, over 80 years old). METHODS: We performed a multicenter retrospective study based on the electronic ICU (eICU) Collaborative Research Database (eICU-CRD), Medical Information Mart for Intensive Care Database (MIMIC-III CareVue and MIMIC-IV), and the Amsterdam University Medical Centers Database (AmsterdamUMCdb). Least Absolute Shrinkage and Selection Operator regression was applied to variables selection. The logistic regression algorithm was used to develop the risk score and a nomogram was further generated to explain the score. RESULTS: We analyzed 23 704 VOPs, including 3 726 deaths (10 183 [13.5% mortality] from eICU-CRD [development set], 12 703 [17.2%] from the MIMIC, and 818 [20.8%] from the AmsterdamUMC [external validation sets]). Thirty-four variables were extracted on the first day of ICU admission, and 10 variables were finally chosen including Glasgow Coma Scale, shock index, respiratory rate, partial pressure of carbon dioxide, lactate, mechanical ventilation (yes vs no), oxygen saturation, Charlson Comorbidity Index, blood urea nitrogen, and urine output. The nomogram was developed based on the 10 variables (area under the receiver operating characteristic curve: training of 0.792, testing of 0.788, MIMIC of 0.764, and AmsterdamUMC of 0.808 [external validating]), which consistently outperformed the Sequential Organ Failure Assessment, acute physiology score III, and simplified acute physiology score II. CONCLUSIONS: We developed and externally validated a nomogram for predicting mortality in VOPs based on 10 commonly measured variables on the first day of ICU admission. It could be a useful tool for clinicians to identify potentially high risks of VOPs.

[Research on procedure to assure the continued validity of the verification results of standard mercury-in-glass thermometer (grade I)].

Standard mercury-in-glass thermometer (Grade I) can be used to perform traceable measurements. Full recalibration is necessary to assure the continued validity of the verification results of the standard mercury-in-glass thermometer (Grade I). The present paper shows researches on procedure for the recalibration at the ice point of standard mercury-in-glass thermometer (Grade I), and points out different calculation ways of the true temperature of the thermostatic bath. The different values of scale correction and adjusted scale correction are compared in this paper.

[Research in filter functionality settings of digital electrocardiograph].

Filter pass-band settings have impact not only on ECG output amplitude, but also on output signal wave-form of some types of digital electrocardiograph. Lower cut-off frequency is decided by Wander filter setting for some types of digital electrocardiograph, and higher cut-off frequency is decided by muscle filter when muscle filter functionality is "on". We research into various filter settings' impact on the output of digital electrocardiograph and have discussions on the malfunctions found in digital electrocardiograph measurement.

[Big data in emergency medcine and Datathon event].

OBJECTIVE: A detailed, high-scale clinical data can be generated in the process of diagnosis and treatment of emergency critically ill patients. The integration and analysis and utilization of these data are of great value for improving the treatment level and efficiency and developing the data-driven clinical assistant decision support. China has large volume of health information resources, however, the construction of healthcare databases and subsequent secondary analysis has just started. With the effort of the Chinese PLA General Hospital in building an emergency database and promoting data sharing, the first emergency database was published in China and a health Datathon was organized utilizing this database, providing experience for clinical data integration, database construction, cross-disciplinary collaboration and data sharing. Referring to the development at home and abroad, this review discussed work in this area and further proposed establishing a big data cooperation for emergency medicine and building a learning healthcare system to integrate more clinical resources and form a closed loop of "clinical database construction-analysis-applications", and enhance the effectiveness of medical big data in reducing medical costs and improving healthcare delivery.

[A cross-disciplinary collaborative "Datathon" model to promote the application of medical big data].

OBJECTIVE: Medical practice generates and stores immense amounts of clinical process data, while integrating and utilization of these data requires interdisciplinary cooperation together with novel models and methods to further promote applications of medical big data and research of artificial intelligence. A "Datathon" model is a novel event of data analysis and is typically organized as intense, short-duration, competitions in which participants with various knowledge and skills cooperate to address clinical questions based on "real world" data. This article introduces the origin of Datathon, organization of the events and relevant practice. The Datathon approach provides innovative solutions to promote cross-disciplinary collaboration and new methods for conducting research of big data in healthcare. It also offers insight into teaming up multi-expertise experts to investigate relevant clinical questions and further accelerate the application of medical big data.

[Analysis of diseases distribution in Medical Information Mart for Intensive Care III database].

OBJECTIVE: To study the distribution of diseases in Medical Information Mart for Intensive Care (MIMIC-III) database in order to provide reference for clinicians and engineers who use MIMIC-III database to solve clinical research problems. METHODS: The exploratory data analysis technologies were used to explore the distribution characteristics of diseases and emergencies of patients (excluding newborns) in MIMIC-III database were explored; then, neonatal gestational age, weight, length of hospital stay in intensive care unit (ICU) were analyzed with the same method. RESULTS: In the MIMIC-III database, 46 428 patients were admitted for the first time, and 49 214 ICU records were recorded. There were 26 076 males and 20 352 females; the median age was 60.5 (38.6, 75.6) years, and most patients were between 60 and 80 years old. The first diagnosis in the disease spectrum analysis was firstly ranked by circulatory diseases (32%), followed by injury and poisoning (14%), digestive system disease (8%), tumor (7%), respiratory disease (6%) and so on. Patients with ischemic heart disease accounted for the largest proportion of circulatory disease (42%), the proportion of these patients gradually increased with age of 60-70 years old, then decreased. However, the proportion of patients with cerebrovascular disease declined first and then increased with age, which was the main cause of death of circulatory system disease (ICU mortality was 22.5%). Injury and poisoning patients showed a significant decrease with age. Digestive system diseases were younger than the general population (most people aged between 50 to 60 years), and non-infectious enteritis and colitis were the main causes of death (ICU mortality was 18.3%). Respiratory infections were predominant in infected patients (34%), but circulatory system infections were the main cause of death (ICU mortality was 25.6%). Secondly, in the neonatal care unit, premature infants accounted for the vast majority (82%). As the gestational age increased, the duration of ICU was decreased, and the mortality was decreased. CONCLUSIONS: The diseases distribution of patients can be provided by MIMIC-III database, which helps to grasp the overview of the volume and age distribution of the target patients in advance, and carry out the next step of research. Meanwhile, it points out the important role of exploratory data analysis in electronic health records analysis.

[Pilot research: construction of emergency rescue database].

OBJECTIVE: To construct a database containing multiple kinds of diseases that can provide "real world" data for first-aid clinical research. METHODS: Structured or non-structured information from hospital information system, laboratory information system, emergency medical system, emergency nursing system and bedside monitoring instruments of patients who visited department of emergency in PLA General Hospital from January 2014 to January 2018 were extracted. Database was created by forms, code writing, and data process. RESULTS: Emergency Rescue Database is a single center database established by PLA General Hospital. The information was collected from the patients who had visited the emergency department in PLA General Hospital since January 2014 to January 2018. The database included 530 585 patients' information of triage and 22 941 patients' information of treatment in critical rescue room, including information related to human demography, triage, medical records, vital signs, lab tests, image and biological examinations and so on. There were 12 tables (PATIENTS, TRIAGE_PATIENTS, EMG_PATIENTS_VISIT, VITAL_SIGNS, CHARTEVENTS, MEDICAL_ORDER, MEDICAL_RECORD, NURSING_RECORD, LAB_TEST_MASTER, LAB_RESULT, MEDICAL_EXAMINATION, EMG_INOUT_RECORD) that containing different kinds of patients' information. CONCLUSIONS: The setup of high quality emergency databases lay solid ground for scientific researches based on data. The model of constructing Emergency Rescue Database could be the reference for other medical institutions to build multiple-diseases databases.