Brain biopsy is commonly employed for the histological diagnosis of complex intracranial diseases. To improve the positive diagnostic rate, the precision of intraoperative tissue sampling is critical. This study evaluated the accuracy of fluorescence imaging technology in rapidly distinguishing tumours from nontumour tissue during surgery, thus providing real-time feedback to surgeons and optimizing the surgical workflow. Biopsy samples from 65 patients were selected for this study. The lesion tissues were sequentially stained with sodium fluorescein and methylene blue, followed by fluorescence imaging via a handheld EndoScell scanner under an intraoperative cellular microscope. Frozen section examinations and haematoxylin-eosin (HE) staining were performed on the same lesion tissue by the pathology department. The time required for fluorescence imaging and pathology of frozen sections was recorded. The results of fluorescence imaging (whether the tissue was a tumour or nontumour tissue) and frozen pathology (whether the tissue was a tumour or nontumour tissue) were also recorded. The HE staining results were used as the final gold standard for diagnosis. The sensitivity, specificity, area under the curve (AUC), Kappa consistency test, and diagnostic efficiency of both methods were calculated. Lesion tissue and diagnostic results were successfully obtained from all 65 patients. When HE-stained histopathology was used as the gold standard, the sensitivity of fluorescence imaging was 100% (95% CI: 0.917-1.000), and the specificity was 63.6% (95% CI: 0.316-0.876). In comparison, the sensitivity of frozen section pathology was 88.9% (95% CI: 0.767-0.954), and the specificity was 100% (95% CI: 0.679-1.000). Both methods demonstrated high diagnostic accuracy. ROC curve analysis revealed that the AUCs for fluorescence imaging and frozen pathology were 0.818 and 0.944, respectively, with no significant difference observed in diagnostic performance (Z = 1.597, P > 0.05). Kappa consistency tests indicated that the Kappa value for frozen pathology compared with HE staining was 0.730 (P < 0.001); for fluorescence imaging compared with HE staining, the Kappa value was 0.744 (P < 0.001), thus demonstrating strong agreement with the HE staining results for both methods. In terms of time efficiency, fluorescence imaging was significantly faster than frozen section pathology [6 (4, 7) min vs. 48 (46, 55) min, Z=-9.856, P < 0.001], thus showing a clear advantage regarding time efficiency for fluorescence imaging. Intraoperative fluorescence imaging via an EndoScell scanner, which represents a novel method for histopathological diagnosis, has high diagnostic accuracy and efficiency. This method provides real-time guidance for tissue sampling strategies in brain biopsy, thereby improving the positive diagnostic rate and reducing surgical risk.