IntroductionFluorescence-guided surgery is one of the rapidly emerging methods of surgical “theranostics.” In this review, we summarize current fluorescence techniques used in neurosurgical practice for brain tumor patients as well as future applications of recent laboratory and translational studies.MethodsReview of the literature.ResultsA wide spectrum of fluorophores that have been tested for brain surgery is reviewed. Beginning with a fluorescein sodium application in 1948 by Moore, fluorescence-guided brain tumor surgery is either routinely applied in some centers or is under active study in clinical trials. Besides the trinity of commonly used drugs (fluorescein sodium, 5-aminolevulinic acid, and indocyanine green), less studied fluorescent stains, such as tetracyclines, cancer-selective alkylphosphocholine analogs, cresyl violet, acridine orange, and acriflavine, can be used for rapid tumor detection and pathological tissue examination. Other emerging agents, such as activity-based probes and targeted molecular probes that can provide biomolecular specificity for surgical visualization and treatment, are reviewed. Furthermore, we review available engineering and optical solutions for fluorescent surgical visualization. Instruments for fluorescent-guided surgery are divided into wide-field imaging systems and hand-held probes. Recent advancements in quantitative fluorescence-guided surgery are discussed.ConclusionWe are standing on the threshold of the era of marker-assisted tumor management. Innovations in the fields of surgical optics, computer image analysis, and molecular bioengineering are advancing fluorescence-guided tumor resection paradigms, leading to cell-level approaches to visualization and resection of brain tumors.

Confocal laser endomicroscopy (CLE) is an advanced optical fluorescence technology undergoing assessment for applications in brain tumor surgery. Many of the CLE images can be distorted and interpreted as nondiagnostic. However, just one neat CLE image might suffice for intraoperative diagnosis of the tumor.While manual examination of thousands of nondiagnostic images during surgery would be impractical, this creates an opportunity for a model to select diagnostic images for the pathologists or surgeons review. In this study, we sought to develop a deep learning model to automatically detect the diagnostic images.We explored the effect of training regimes and ensemble modeling and localized histological features from diagnostic CLE images. The developed model could achieve higher agreement with the ground truth than the other human observers.With the speed and precision of the proposed method, it has potential to be integrated into the operative workflow in the brain tumor surgery.