Research & Innovation

Artificial Intelligence in Medicine

:::

Intelligent Sepsis Early Prediction System

iSEPS Intelligent Bacteremia risk Assessment with Complete blood count and Cell population data

The iSEPS system (Intelligent Bacteremia risk Assessment with Complete blood count and Cell population data) using artificial intelligence analysis techniques to provide the risk of bacteremia within an hour. The iSEPS utilizes data obtained from new generation blood analyzer, including complete blood count(CBC) tests as well as 56 cell population data generated during the testing process. Alongside routine CBC reports, it provides assessment of bacteremia risk simultaneously. This can appropriately alert clinicians to atypical presentations of bacteremia, prompt infection assessment and administer necessary treatment. The implementation of iSEPS can enhance the timeliness of infectious disease management, achieving the goal of real-time assessment and precise treatment.

The introduction of iSEPS provides physicians with a background guardian against bacteremia. The iSEPS system requires neither additional blood sample nor extra clinical work from physicians, nurses, or medical technologists. The iSEPS system utilizes artificial intelligence to integrate data and infer the bacteremia risk for each case. Through an information streaming platform, iSEPS synchronizes bacteremia risk assessment while physicians perform routine CBC tests. Leveraging artificial intelligence technology, iSEPS offers innovative, real-time, precise, and clinically relevant information within traditional examinations for healthcare providers.

▲ The iSEPS utilizes intelligence technology technology to provide bacteremia risk assessment within one hour, compared to traditional blood cultures, it can predict bacteremia risk 18-70 hours in advance.

▲ The iSEPS system has been fully implemented at China Medical University Hospital. The accuracy of the iSEPS system remained high at 0.844 after full implementation. It comprehensively provides trustworthy bacteremia risk assessments.

Journal Article

  • Machine learning of cell population data, complete blood count, and differential count parameters for early prediction of bacteremia among adult patients with suspected bacterial infections and blood culture sampling in emergency departments.
    https://pubmed.ncbi.nlm.nih.gov/37244761/

Awards

2024

  • National Innovation Award&Future Tech Award
  • Symbol of National Quality

2023

  • National Healthcare Quality Award (bronze award)


Contact window

Emergency department
Phone: +886-4-22052121 ext 15433
Email: 028761@tool.caaumed.org.tw

Stay connected with CMUH
How to get to CMUH the map of hospital