News Release

New LED-based imaging system could transform cancer detection in endoscopy

LED-based hyperspectral imaging system shows promise for cancer detection during endoscopy

Peer-Reviewed Publication

SPIE--International Society for Optics and Photonics

The proposed hyperspectral endoscopic imaging system includes a spectral LED array in the catheter tip.

image: 

The proposed hyperspectral endoscopic imaging system includes a spectral LED array in the catheter tip. 

view more 

Credit: N. Modir et al., doi 10.1117/1.JMI.12.3.035002.

Gastrointestinal cancers remain among the most common forms of cancer. While endoscopy has become a cornerstone of cancer screening and diagnosis over the past two decades, the procedure still misses approximately 8 to 11 percent of tumors, due to visibility limitations. Now, researchers have developed a prototype imaging system that could significantly improve doctors' ability to detect cancerous tissue during endoscopic procedures.

The new approach, reported in the Journal of Medical Imaging, combines light-emitting diodes (LEDs) with hyperspectral imaging technology to create detailed maps of tissue properties that are invisible to conventional endoscopic cameras. Unlike standard endoscopy, which captures images using broad red, green, and blue color channels, hyperspectral imaging records data across numerous narrow wavelength bands, including light beyond the visible spectrum. This allows the system to detect biochemical changes in cancerous tissue that produce distinct spectral signatures.

Testing the LED Array Concept

The research team led by Dr. Baowei Fei, Professor and Cecil H. and Ida Green Chair in Systems Biology Science at the University of Texas at Dallas (UT Dallas) Quantitative BioImaging Laboratory, designed and tested a prototype system built around an array of 18 LEDs, each emitting light at different wavelengths ranging from 405 nanometers to 910 nanometers. The system uses a monochrome camera to capture images as each LED illuminates the target tissue in sequence, building up a complete hyperspectral dataset.

The researchers evaluated their LED-based system by imaging both normal and cancerous tissue samples removed during surgery. They investigated how different imaging conditions affected the quality of the hyperspectral data and compared their results with those from a reference hyperspectral camera system used as a gold standard.

Promising Results for Medical Applications

The LED-based prototype successfully captured hyperspectral signatures from various tissue types, producing data comparable to the reference system. The researchers found that their approach could achieve imaging rates of over 10 hyperspectral datasets per second, which approaches the real-time speeds needed for practical endoscopic procedures.

"Our research shows the feasibility of employing a spectral LED array as the illumination source for high-speed and high-quality hyperspectral imaging," commented Naeeme Modir, PhD candidate at UT Dallas and first author. “Our findings suggest that LED-based systems could open new possibilities for hyperspectral imaging applications,” she remarked.

Advantages Over Current Technology

The LED-based approach offers several potential benefits over existing hyperspectral endoscopy systems. Traditional hyperspectral endoscopes often require fiber optic bundles to deliver light through the working channel of the endoscope, which limits the available space for other medical instruments. The new design places the LEDs directly at the tip of the endoscope, leaving the working channel free for surgical tools or other procedures.

The system also allows doctors to adjust the intensity of each LED individually based on the distance to the target tissue. This real-time control helps prevent problems like sensor saturation from excessive light or noisy images from insufficient illumination. Additionally, because tissue is exposed to each wavelength for only a brief moment, the approach may reduce potential damage from prolonged light exposure compared to systems that use broad-spectrum white light.

Technical Innovation and Future Applications

The researchers designed their system using wavelength scanning, where the LEDs illuminate the tissue sequentially at different wavelengths. This method balances the need for speed with the requirement for high spectral resolution. The approach contrasts with other hyperspectral imaging methods that either sacrifice speed for resolution or resolution for speed.

Micro-LED technology makes the system practical for medical use. These microscopic LEDs, with footprints smaller than 400 micrometers square, can be mounted on a circuit board surrounding the camera at the endoscope tip. This compact design enables the integration of dozens of LEDs without significantly increasing the size of the endoscope.

The ability to customize which spectral bands are used for imaging represents another advantage. Doctors could potentially select specific wavelength combinations optimized for detecting particular types of cancer or other tissue abnormalities, improving diagnostic accuracy while maintaining fast imaging speeds.

Path Forward

“While the prototype demonstrates the technical feasibility of LED-based hyperspectral endoscopy, additional development will be needed before the technology reaches clinical practice,” notes Modir. “We established that this approach can produce high-quality hyperspectral data at practical imaging speeds—hopefully providing a foundation for future medical applications.”

The study represents a significant step toward more effective endoscopic cancer screening. By potentially reducing the number of missed tumors and enabling real-time tissue assessment over larger areas, LED-based hyperspectral imaging could help doctors make faster, more accurate diagnoses while reducing the need for unnecessary tissue removal and pathological testing.

For details, see the original Gold Open Access article by N. Modir et al., “LED-based, real-time, hyperspectral imaging device,” J. Med. Imaging 12(3), 035002 (2025), doi: 10.1117/1.JMI.12.3.035002.


Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.