Microscope-integrated optical coherence tomography for in vivo human brain tumor detection with artificial...

Area of Science:

• Neurosurgery
• Medical Imaging
• Biomedical Engineering

Background:

• Optical coherence tomography (OCT) has potential for intraoperative brain tumor margin diagnostics.
• Limited in vivo human data exists on OCT's accuracy for residual brain tumor detection (RTD).

Purpose of the Study:

• To evaluate the in vivo feasibility and accuracy of a microscope-integrated OCT system for RTD.
• To assess automated scan analysis for intraoperative tissue classification.

Main Methods:

• 3D OCT scanning of resection edges in 18 brain tumor patients.
• Machine learning and artificial intelligence (AI) methods for scan classification.
• Comparison of OCT properties with neuropathologist-confirmed biopsy ground truth.

Main Results:

• In vivo OCT scanning is feasible and integrates into surgical workflow.
• OCT effectively distinguished white matter from tumor infiltration (85% accuracy with SVM).
• AI approaches (neural network, autoencoder) achieved 82-85% accuracy for diseased brain detection.

Conclusions:

• In vivo OCT scanning provides significant value for intraoperative RTD in brain tumors.
• OCT can complement existing methods for guiding surgical resection decisions.
• Further research can enhance intraoperative brain tumor margin assessment.