Texture Analysis of Fibrous Meningioma Using Label-Free Multiphoton Microscopy.

Abstract

Fibrous meningiomas, a common type of brain tumor, present surgical challenges due to their variable hardness, which is crucial for complete resection and patient prognosis. This study explores the use of label-free multiphoton microscopy (MPM) for the objective assessment of the texture of fibrous meningiomas. Fresh tumor samples from 20 patients were analyzed using both multichannel and lambda mode MPM, with quantitative image analysis algorithms determining collagen content and multi-peak spectral fitting providing additional optical collagen metrics. The study compared medium and hard fibrous meningiomas, utilizing receiver operating characteristic analysis to evaluate predictive performance. Microstructural features were clearly visualized, enabling accurate diagnosis. Collagen-related parameters significantly differentiated between moderate and hard tumors (p < 0.05). High predictive values were observed for collagen content, collagen-to-NADH-free ratio, and collagen-to-FAD ratio (AUC = 0.748-0.839). A multivariate logistic model combining these biomarkers significantly improved diagnostic accuracy (AUC = 0.907). The findings suggest that MPM, with its ability to visualize and quantify microstructures such as collagen and cells without the need for staining, holds strong potential for rapid, objective, and accurate assessment of tumor texture during neurosurgery. The integration of MPM with multiphoton endoscopy paves the way for potential in vivo applications in the future.