User Settings
Article

Estimation of blood flow in the upper gastrointestinal tract by analysis of endoscopic true color images.

R S Jacoby,Franz Maximilian Zwiebel,Rainer Herpers,Karl‐Hans Englmeier-1995-01-01-PubMed
0

TL;DRAbstract

A method is presented that estimates the local blood flow in the mucosa of the organs of the upper gastrointestinal tract; the method is based on the analysis of endoscopic true-color images. The quantity of blood flow is approximated by the estimation of the hemoglobin concentration in the mucosa. The first step of our algorithm consists of a neural segmentation, which excludes artifacts of the images that interfere with further computation. Next, a transformation of the image data is performed within the RGB-color space in order to obtain an estimation of the blood distribution, which is independent of the local brightness in the images. Finally the quantity of blood flow is estimated on the basis of physical laws of reflectance spectroscopy. Our method is characterized by the following features: 1) It computes an estimation of the blood flow for a whole endoscopic image; as such it is more powerful than local measuring methods; 2) Our method does not need any modifications of the en

Chat with Paper

AI Agents for this Paper

A method is presented that estimates the local blood flow in the mucosa of the organs of the upper gastrointestinal tract; the method is based on the analysis of endoscopic true-color images. The quantity of blood flow is approximated by the estimation of the hemoglobin concentration in the mucosa. The first step of our algorithm consists of a neural segmentation, which excludes artifacts of the images that interfere with further computation. Next, a transformation of the image data is performed within the RGB-color space in order to obtain an estimation of the blood distribution, which is independent of the local brightness in the images. Finally the quantity of blood flow is estimated on the basis of physical laws of reflectance spectroscopy. Our method is characterized by the following features: 1) It computes an estimation of the blood flow for a whole endoscopic image; as such it is more powerful than local measuring methods; 2) Our method does not need any modifications of the en

Keywords

Blood flowBrightnessRGB color modelComputer visionArtificial intelligenceColor spaceComputationComputer science

Chat

Click to start Chat