A full-spectrum analysis method for monitoring blood oxygen saturation and blood cell accumulation in an extracorporeal circulation pipeline

By combining full-spectrum analysis and deep learning, and utilizing a hybrid sensor composed of a CMOS image sensor and a photodiode, the problems of large errors and invasiveness in monitoring blood oxygen saturation and hematocrit in extracorporeal circulation tubing have been solved, achieving high-precision, low-cost non-invasive monitoring.

CN122141041APending Publication Date: 2026-06-05YIXING PEOPLES HOSPITAL +1
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
YIXING PEOPLES HOSPITAL
Filing Date
2026-01-15
Publication Date
2026-06-05

AI Technical Summary

Technical Problem

Existing methods for monitoring blood oxygen saturation and hematocrit in extracorporeal circulation tubing have large errors under pulseless flow conditions, and changes in hematocrit interfere with the measurement. Traditional equipment is also expensive and highly invasive.

Method used

Employing a full-spectrum analysis method, this system utilizes a hybrid sensor composed of a CMOS image sensor and a photodiode, combined with a deep learning model. By simultaneously acquiring visible and infrared light, it extracts characteristic information on blood oxygen saturation and hematocrit, achieving non-invasive and low-cost monitoring.

Benefits of technology

It enables high-precision monitoring of blood oxygen saturation and hematocrit under pulseless flow conditions, reduces equipment costs, avoids invasive risks, and effectively decouples HCT interference, thus improving the robustness of the measurement.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122141041A_ABST
    Figure CN122141041A_ABST
Patent Text Reader

Abstract

The application discloses a full-spectrum analysis method for monitoring blood oxygen saturation and hematocrit in an extracorporeal circulation pipeline, and comprises the following steps: step 1, constructing a photoelectric acquisition device; step 2, performing system initialization calibration; acquiring CMOS image sensor data, photodiode sensor dark current calibration data and empty pipeline reference data; step 3, using a main control unit to control a mixed light source module and a mixed sensor module to perform time sequence synchronous acquisition; acquiring a transmission image of a visible light band and a transmission light intensity voltage of an infrared band at a fixed frame rate; step 4, pre-processing the acquired transmission image data and transmission light intensity voltage data to generate an optical characteristic map reflecting the pipeline and an optical density vector reflecting absolute light absorption; and step 5, inputting the pre-processed optical characteristic map and optical density vector into a pre-trained double-flow deep neural network model to obtain quantitative values of blood oxygen saturation and hematocrit of liquid in the pipeline.
Need to check novelty before this filing date? Find Prior Art