Client Backstory

The client is a leading health network and was looking to improve the quality of and reduce the readmission rate of the patient.

Help physicians treat their patients by monitoring & capturing patients’ biometric data in real-time and analyzing it to improve the quality of care along with reducing the likelihood of patient readmissions.

Challenges

  • Capturing real-time data from the ICU’s IoT Devices
  • Storing of biometric data at high speed
  • Poor visibility on the real-time patient condition
  • Inability to identify the ideal treatment for better outcomes

Strategic Approach

Set up Azure IoT Hub for capturing real-time data from the ICU IoT devices.

Connected Azure IoT Hub to Azure Stream Analytics. Set up window-creation functions for the ICU data. These functions helped to aggregate the data for each window. At the same time, set up the IoT Hub to move the streaming data to Azure Data Lake Storage by using Azure Functions.

We further set up Azure Functions to store the Azure Stream Analytics aggregates in Azure Data Lake Storage Gen2.

Used Azure Data Factory to load data from the data lake into Azure Synapse Analytics which supported the chief medical officer's needs. After the data was loaded, transformations occurred within Azure Synapse Analytics.

Parallelly we connected the Azure Machine Learning service to Azure Data Lake Storage to conduct predictive analytics.

Connected Power BI to Stream Analytics so that we can fetch the real-time aggregates for the patient data. We also connected Azure Synapse Analytics to pull the historical data and created a comprehensive dashboard.

Impact

  • Reduced the readmission of the patient by 2%.
  • In a shorter period of time, the patient condition was observed to be improved by 4%.
  • The system helped our client to use the ICU ward optimally and helped the client to cater to more patients efficiently.
  • Elevated the brand identity and growth manifolds.

Not sure how to capitalize Machine Learning in your business?