Clinical research on a microscope-integrated system paves the way for employing OCT to outline tumor boundaries and uncover hidden brain structures.
Researchers have effectively combined a high-speed optical coherence tomography (MHz-OCT) system with an available neurosurgical microscope, showcasing its practical application in clinical settings. This breakthrough marks a significant advancement toward creating an OCT tool that could help delineate tumor boundaries during brain surgeries.
OCT is a non-invasive imaging method that delivers detailed cross-sectional views of tissues, allowing for visualization of deeper structures. While this technique is extensively utilized in fields like ophthalmology and cardiology, most existing commercial OCT systems can only capture about 30 two-dimensional images per second.
“Our MHz-OCT system operates at exceptional speeds, approximately 20 times quicker than the typical OCT systems,” said Wolfgang Draxinger from Universität zu Lübeck. “This speed enables it to generate three-dimensional images that can penetrate below the brain’s surface, potentially processed with AI to identify unhealthy areas that might remain concealed with other imaging technologies.”
In a publication from the Optica Publishing Group journal Biomedical Optics Express, a team led by Robert Huber discussed findings from a clinical trial of the MHz-OCT system integrated into microscopes. They demonstrated that the system can produce high-quality volumetric OCT scans during surgery in seconds, with images ready for immediate analysis.
“We envision our integrated MHz-OCT system being utilized not only for brain tumor surgeries but across all neurosurgical procedures, as it can capture high-contrast images of anatomical features like blood vessels even through the protective membrane covering the brain,” noted Draxinger, the lead author of the new study. “This could greatly enhance outcomes for surgeries that require precise information about structures beneath the brain’s surface, like deep brain stimulation for Parkinson’s disease.”
Enhancing OCT Speed
The researchers have focused on accelerating OCT technology by optimizing the light sources and sensors used and creating software capable of handling the vast data produced. This has led to the creation of an MHz-OCT system that can conduct over a million depth scans per second.
This rapid scanning capability is made possible due to the integration of a Fourier domain mode locking laser, initially conceived by Huber during his PhD at MIT in 2005, with the help of his advisors James G. Fujimoto, Eric Swanson, and David Huang—all pioneers of OCT. Furthermore, advancements in graphics processing unit (GPU) technology in recent years have provided the necessary computational power to transform raw OCT signals into manageable images without requiring bulky computing equipment.
To evaluate whether the MHz-OCT system could effectively visualize brain tumor margins, researchers combined it with a specialized microscope already in use by surgeons for enhanced views of the brain.
Testing in the Operating Room
After assembling the integrated system, they began testing it with calibration targets and tissue-like phantoms. Once satisfied with these initial results, they moved on to patient safety trials and subsequently conducted a clinical study involving 30 patients undergoing brain tumor removal surgery.
“We discovered that our system fits seamlessly into the standard operating room workflow, facing no significant technological obstacles,” stated Draxinger. “The image quality exceeded our initial expectations, which were conservative since we were modifying an existing system.”
Throughout the clinical study, the researchers accumulated around 10 TB of OCT imaging data alongside corresponding pathological histology information. They indicate that they are still in the preliminary phases of deciphering the data generated by the new system and formulating AI techniques for tissue classification. Therefore, it may take several years before this technology can be broadly adopted in brain tumor removal surgeries.
Additionally, they are planning a new study to use the system to identify the precise locations of brain activity in response to external stimuli during neurosurgery. This could significantly enhance the accuracy of placing neuroprosthetic electrodes, thereby improving the control of prosthetic devices through the brain’s electrical signals.
