5 March 2018 Brain tissue analysis using texture features based on optical coherence tomography images
Author Affiliations +
Brain tissue differentiation is highly demanded in neurosurgeries, i.e. tumor resection. Exact navigation during the surgery is essential in order to guarantee best life quality afterwards. So far, no suitable method has been found that perfectly covers this demands. With optical coherence tomography (OCT), fast three dimensional images can be obtained in vivo and contactless with a resolution of 1-15 μm. With these specifications OCT is a promising tool to support neurosurgeries. Here, we investigate ex vivo samples of meningioma, healthy white and healthy gray matter in a preliminary study towards in vivo brain tumor removal assistance. Raw OCT images already display structural variations for different tissue types, especially meningioma. But, in order to achieve neurosurgical guidance directly during resection, an automated differentiation approach is desired. For this reason, we employ different texture feature based algorithms, perform a Principal Component Analysis afterwards and then train a Support Vector Machine classifier. In the future we will try different combinations of texture features and perform in vivo measurements in order to validate our findings.
© (2018) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Marcel Lenz, Marcel Lenz, Robin Krug, Robin Krug, Christopher Dillmann, Christopher Dillmann, Nils C. Gerhardt, Nils C. Gerhardt, Hubert Welp, Hubert Welp, Kirsten Schmieder, Kirsten Schmieder, Martin R. Hofmann, Martin R. Hofmann, "Brain tissue analysis using texture features based on optical coherence tomography images", Proc. SPIE 10483, Optical Coherence Tomography and Coherence Domain Optical Methods in Biomedicine XXII, 104833B (5 March 2018); doi: 10.1117/12.2292032; https://doi.org/10.1117/12.2292032

Back to Top