Intraoperative In Vivo Cross-Polarization Optical Coherence Tomography Dataset of Human Brain Gliomas

This dataset was used for the following publication: Radik Zinatullin, Alexander Sovetsky, Artem Grishin, Elena Kiseleva, Liudmila Kukhnina, Svetlana Korikova, Alexander Matveyev, Vladimir Zaitsev, Konstantin Yashin, and Lev Matveev. "Human-in-the-Loop Enhances Machine Learning Inference in Intraoperative Optical Coherence Tomography Glioma Imaging." Medical Sciences (2026) [accepted].This dataset contains raw intraoperative in vivo cross-polarization optical coherence tomography (CP-OCT) scans of human brain tissue. The data was collected in vivo from 27 patients during surgical resection of malignant brain gliomas. All scanned tissue samples were precisely excised and submitted for histopathological examination to establish the ground truth. The dataset includes 86 structural scans in total, provided in a .zip archive (Brain_CPOCT_Scans_Raw.zip). The data is divided into two folders (classes) based on histological confirmation:Class 1 (Non-malignant): 44 scans of intact white matter and non-malignant damaged white matter.Class 2 (Malignant): 42 scans of tumor tissue (including grade I-III gliomas, grade IV IDH-mutant astrocytoma, and glioblastoma). Image Specifications: Each provided .png file represents a combined B-scan containing two co-registered optical channels.Overall Image Dimensions: 400 pixels (lateral) × 512 pixels (axial/vertical).Cross-polarization channel: Located in the top half of the image (pixels 1 to 256 vertically).Co-polarization channel: Located in the bottom half of the image (pixels 257 to 512 vertically). Therefore, each individual scan channel is 400 pixels in width (laterally) and 256 pixels in depth. Acquisition Device: The structural scans were acquired directly in the operating room using a time-domain polarization-sensitive optical coherence tomography system (OCT-1300U, BioMedTech LLC, Nizhny Novgorod, Russia). The scanner operates at a central wavelength of 1300 nm. ----------- Terms of Use & Citation: This dataset is published under an open-access license. However, if you use this dataset in your research, software development, or any other related work, you must cite the following original publication:Radik Zinatullin, Alexander Sovetsky, Artem Grishin, Elena Kiseleva, Liudmila Kukhnina, Svetlana Korikova, Alexander Matveyev, Vladimir Zaitsev, Konstantin Yashin, and Lev Matveev. "Human-in-the-Loop Enhances Machine Learning Inference in Intraoperative Optical Coherence Tomography Glioma Imaging." Medical Sciences (2026) [accepted].----------- This dataset was curated and published with the support of the Russian Science Foundation (RSF):RSF Grant No. 23-75-10068: "The metabolic and structural features of tumor and peritumoral zone in glioblastoma based on multimodal visualization using MRI, OCT and FLIM" (https://rscf.ru/en/project/23-75-10068/) - specifically for data aquisition and histopathological examination;RSF Grant № 25-12-20032: "New Approaches to the Development of Algorithms for Analyzing OCT Scans: Modification and Optimization of Large Models Based on Physical Principles and Conditions of OCT Signal Formation" (https://rscf.ru/en/project/25-12-20032/) - specifically for data curation for models training and benchmarking tools.