Development of a Tissue Oxygen Saturation Imaging Technique Based on Electronic Endoscopy

Hypoxemia, a common clinical phenomenon, is closely associated with various pathological conditions, including tumors and inflammatory diseases. The detection of tissue hypoxia holds significant potential for the detailed diagnosis of these conditions. Existing endoscopes, with the aid of probes, can perform point measurements of tissue oxygen saturation (<i>S</i><i>t</i><i>O</i>₂), a well-established method for assessment. However, this approach has limitations, including a small measurement range and slow measurement speed, which restrict its further application. In addition, hyperspectral imaging featured with a large number of spectral bands is used in conventional strategies, which increases the complexity of the endoscope. This work aims to develop an endoscopic technique for measuring <i>S</i><i>t</i><i>O</i>₂ in digestive tract tissue, ensuring compatibility with existing imaging systems, using fewer spectral bands, and without requiring significant modifications to current surgical and diagnostic endoscopic systems. The Monte Carlo method is used to numerically model light transport and its interaction with human tissues, establishing the relationship between tissue reflectance spectra and tissue properties. Based on this, a lookup-table-based (LUT) model has been developed to enable hardware acceleration for <i>S</i><i>t</i><i>O</i>₂ measurement. Three illumination wavelength bands from commercially available LEDs were selected to analyze <i>S</i><i>t</i><i>O</i>₂ from tissue reflectance spectra. The effectiveness and accuracy of this method were validated using a prototype device, with experiments conducted on tissue-simulating phantoms. The experimental results are in good agreement with the measurements from a visible light spectrophotometer, with the MAE of tissue oxygen values no more than 4.4%. These findings confirm the effectiveness of this technology in assessing <i>S</i><i>t</i><i>O</i>₂ in digestive tract tissue and demonstrate its high potential for real-time blood oxygen measurement with a simple system and low-cost characteristics.