This paper investigates the use of hyperspectral imaging (HSI) combined with machine learning for intraoperative brain tumor detection. A robust k-fold cross-validation approach was used to evaluate the performance of HSI in identifying and delineating brain tumors. Analysis of an in-vivo brain database (61 HS images from 34 patients) achieved a median macro F1-Score of 70.2 ± 7.9% on the test set using both spectral and spatial information. The study provides a benchmark for further development of real-time decision support tools in neurosurgery.

Raquel Leon, Himar Fabelo, Samuel Ortega, Ines A. Cruz-Guerrero, Daniel Ulises Campos-Delgado, Adam Szolna, Juan F. Piñeiro, Carlos Espino, Aruma J. O'Shanahan, Maria Hernandez, David Carrera, Sara Bisshopp, Coralia Sosa, Francisco J. Balea-Fernandez, Jesus Morera, Bernardino Clavo, Gustavo M. Callico