Artikel
The neurosurgical benefit of contactless in vivo optical coherence tomography regarding residual tumor detection – a clinical study
Der neurochirurgische Nutzen von berührunglsloser Optischer Kohärenztomographie zur in vivo Erkennung von Residualtumor – eine klinische Studie
Suche in Medline nach
Autoren
-
Patrick Kuppler
- Universitätsklinikum Schleswig-Holstein, Campus Lübeck, Institut für Neurochirurgie, Lübeck, Deutschland
-
Paul Strenge
- Medizinisches Laserzentrum Lübeck GmbH, Lübeck, Deutschland
-
Wolfgang Draxinger
- Universität zu Lübeck, Institut für Biomedizinische Optik, Lübeck, Deutschland
-
Sonja Spahr-Hess
- Universitätsklinikum Schleswig-Holstein, Campus Lübeck, Institut für Neurochirurgie, Lübeck, Deutschland
-
Christian Hagel
- Universitätsklinikum Hamburg-Eppendorf, Institut für Neuropathologie, Hamburg, Deutschland
-
Christin Grill
- Universität zu Lübeck, Institut für Biomedizinische Optik, Lübeck, Deutschland
-
Dirk Theisen-Kunde
- Medizinisches Laserzentrum Lübeck GmbH, Lübeck, Deutschland
-
Ralf Brinkmann
- Medizinisches Laserzentrum Lübeck GmbH, Lübeck, Deutschland; Universität zu Lübeck, Institut für Biomedizinische Optik, Lübeck, Deutschland
-
Robert Huber
- Universität zu Lübeck, Institut für Biomedizinische Optik, Lübeck, Deutschland
-
Birgit Lange
- Medizinisches Laserzentrum Lübeck GmbH, Lübeck, Deutschland
-
Mario Matteo Bonsanto
- Universitätsklinikum Schleswig-Holstein, Campus Lübeck, Institut für Neurochirurgie, Lübeck, Deutschland
| Veröffentlicht: | 25. Mai 2022 |
|---|
Gliederung
Text
Objective: The intraoperative distinction between tumorous brain and healthy brain tissue is still a challenge, especially in the detection of residual tumor at the tumor border during resection. In recent years, several groups have shown that optical coherence tomography (OCT) has the potential to identify tumorous brain tissue with the help of structural and optical image features. However, there is little evidence on human in vivo application of this technology, especially regarding applicability and accuracy of residual tumor detection (RTD). In this study, we investigated the neurosurgical benefit of this system in 21 brain tumor patients.
Methods: In vivo OCT volume scans (C-Scans) were taken with a spectral domain (SD) OCT system integrated into a surgical microscope (Haag-Streit/OptMedt iOCT; Figure 1 [Fig. 1]) within resection cavities and analyzed using a variety of approaches. We established a visual classifier with qualitative OCT image properties (Figure 2 [Fig. 2]) to distinguish healthy brain from tumorous tissue. Further, OCT signal intensity (I) and attenuation (At) were assessed for their distribution and predictive force. Then, we applied deep learning algorithms (DLA), augmenting image recognition through artificial intelligence (AI). The results were compared to histopathology from tissue samples taken from within the same resection cavities and investigated for accuracy of RTD. Also, current standard techniques i.e., fluorescence guided surgery, neuronavigation and early post-operative MRI were compared to our findings.
Results: Analysis of raw in vivo iOCT scans did not contain satisfying results in accuracy of RTD and applicability when compared to already established methods. Analysis of post-processed images and essentially the use of AI increased that accuracy significantly.
Conclusion: Due to still inconvenient surgical application and the need of image post-processing, the use of this iOCT system has not yet shown to improve the intraoperative decision-making process concerning the extend of tumor resection. However, qualitative and quantitative in vivo OCT data analysis has proven to contain additional information on RTD, supporting what has been well described for ex vivo OCT brain tumor scanning. The use of DLA for image feature recognition has shown promising results and might be crucial to achieve high accuracy in tumor border detection that might complement current intraoperative methods and thereby enhance the extend of brain tumor excision.
