Monitoring Framework for Clinical ETL Processes and Associated Performance Resources
Guardado en:
| Udgivet i: | PQDT - Global (2020) |
|---|---|
| Hovedforfatter: | |
| Udgivet: |
ProQuest Dissertations & Theses
|
| Fag: | |
| Online adgang: | Citation/Abstract Full Text - PDF |
| Tags: |
Ingen Tags, Vær først til at tagge denne postø!
|
| Resumen: | The dissertation’s Business Intelligence Architecture is organized in two environments: backend (development environment) and front-end (visualization environment). In the back-end, the process begins with the collection of clinical de-identified data, in CSV files format, from the MIMIC-III demo v.1.4 and subsequent extraction, transformation, and loading (ETL) of the data to the dimensions and facts of the Clinical Data Warehouse (DW).The Clinical ETL process was implemented, on-premises, using Integration Services modules from the Microsoft SQL Server (SS) and, in the Microsoft Azure environment, using the Azure Blob Storage and Data Factory resources.The Business intelligence systems are always associated with Reporting Technologies that create unified, user-friendly dashboards to analyze metrics and key performance indicators. The Clinical Visual Solution was developed in the Power BI Desktop (PBID) and is divided into five data marts: hospital admission services, electronic charted measurements, medical interventions, microbiology, and laboratory tests.Clinical solutions provide quick and effective business decisions for medical and administrative professionals while saving valuable time and resources. In the developed solution, it is possible to evaluate the medical professionals’ workload, the number of medical interventions, electronic charted measurements, microbiology and laboratory tests performed by medical category, date (month, trimester, and year), caregiver, and Intensive Care Unit (ICU). It is also possible to analyze the type of admission, the hospital admission’s survival rate, and the patient’s average medical assistance waiting (time between the admission and the first performed test) and hospitalization time.The clinical DW integrates multiple heterogeneous data sources in the healthcare sector, providing an optimized and effective information platform for health decision-makers. Thus, emerging the necessity to built a central framework to monitor all the Clinical ETL processes, implemented on-premises or in the cloud.The central framework was implemented in the Microsoft Azure Environment to monitor the ETL message errors and resources’ performance metrics (e.g., CPU, RAM) of different Clinical Solutions. The monitoring information (ETL errors and the resources’ performance metrics) were sent via e-mail, using the SendGrid Web API service. The Azure Logic Application interprets the e-mail body and subject to retrieve the monitoring information and store it into the respective Azure SQL database tables. Finally, the end-user can analyze the monitoring dashboard and be alerted if one or more projects fail or are close to failure. |
|---|---|
| ISBN: | 9798382110172 |
| Fuente: | ProQuest Dissertations & Theses Global |