![](/assets/bioengineering/_processed_/5/3/csm_Firefly_Biomedical_Optoacoustic_Sensing__abstract_design__green_and_blue__datatransfer_8787_a505882708.jpg)
AI and multispectral imaging improve safety in keyhole surgery
While conventional devices (right) generate an RGB video image, the newly developed laparoscope (left) uses a multispectral camera. This also makes it possible to visualize functional properties of the tissue.
Laparoscopic (keyhole) surgery has become a staple of cancer diagnosis and therapy. While characterization of the tissue blood flow is crucial in various procedures, such as partial kidney removal, doing so through visual inspection remains highly challenging. Based on a compact and lightweight multispectral camera, we developed a real-time imaging system that allows surgeons to complement the conventional view of the patient with functional information in the shape of a video. To enable contrast agent–free monitoring of the blood flow during laparoscopic kidney removal, we use an artificial intelligence method that only requires the analysis of data from the patient at hand. A clinical trial with patients demonstrates the feasibility of our approach and highlights the potential of spectral imaging combined with advanced deep learning–based analysis tools for fast, efficient, reliable, and safe functional laparoscopic imaging.
Einzelnes Bild
![](/assets/bioengineering/_processed_/9/d/csm_Figure1_343c79f1f5.png)
Our multispectral imaging-based approach enables continuous, contrast agent–free, real-time monitoring of the blood flow in laparoscopic surgery. The clinical state of the art (brown) to verify successful interruption of the blood flow via clamping of arteries is based on fluorescence imaging: Clamped tissue lacks a fluorescent signal, whereas normal tissue fluoresces. Our approach (green) is based on noninvasive and contrast agent–free multispectral imaging at video rate. Invertible neural networks trained on multispectral video sequences before clamping are capable of detecting clamped tissue areas as outliers in real time.
More details:
Publication:
Spectral imaging enables contrast agent–free real-time ischemia monitoring in laparoscopic surgery, Ayala L., Adler T.J., Seidlitz S. et al. Sciences Advances, 2023, DOI: 10.1126/sciadv.add6778
Contact
Lena Maier-Hein
Deutsches Krebsforschungszentrum Heidelberg
E-Mail