Yin Hang, Zhang Xin, Zhao Zheng, Cao Chong, Xu Minhua, Zhou Suhongrui, Xuan Tian, Jin Ziyi, Han Limei, Fan Yang, Wang Cong, Zhu Xiao, Mao Ying, Yu Jinhua, Li Cong
MOE Key Laboratory of Smart Drug Delivery, MOE Innovative Center for New Drug Development of Immune Inflammatory Diseases, School of Pharmacy, Fudan University, Shanghai, 201203, China.
Department of Neurosurgery, Huashan Hospital, Shanghai Medical College, Fudan University, National Center for Neurological Disorders, Shanghai Key Laboratory of Brain Function and Restoration and Neural Regeneration, Neurosurgical Institute of Fudan University, Shanghai Clinical Medical Center of Neurosurgery, Shanghai, 200040, China.
Adv Sci (Weinh). 2025 Jul;12(26):e2503360. doi: 10.1002/advs.202503360. Epub 2025 Apr 2.
Intraoperative identification of the isocitrate dehydrogenase type 1 (IDH1) genotype, a key molecular marker in glioma, is essential for optimizing surgical strategies and tailoring post-surgical treatments. However, current clinical practices lack effective methods for real-time IDH1 genotype detection during surgery. Here, a novel strategy is proposed for intraoperative IDH1 genotype identification by simultaneously measuring two redox-related metabolites. A surface-enhanced Raman scattering (SERS) probe is developed to detect glutathione and hydrogen peroxide concentrations through orthogonally responsive Raman signals. Additionally, a deep learning algorithm is implemented, leveraging 2D Raman spectra transformation and multi-task learning to enhance measurement speed and accuracy. This AI-assisted SERS approach can identify the IDH1 genotype in glioma patients within 7 min. In a cohort of 31 glioma patients, the system achieved an area under the receiver operating characteristic curve of 0.985 for accurate IDH1 genotype differentiation. This study holds significant promise for refining surgical decision-making and personalizing post-surgical treatments by enabling rapid intra-operative molecular biomarker identification.
术中识别异柠檬酸脱氢酶1型（IDH1）基因型是胶质瘤的关键分子标志物，对于优化手术策略和定制术后治疗至关重要。然而，目前的临床实践缺乏在手术期间实时检测IDH1基因型的有效方法。在此，提出了一种通过同时测量两种氧化还原相关代谢物来术中识别IDH1基因型的新策略。开发了一种表面增强拉曼散射（SERS）探针，通过正交响应拉曼信号检测谷胱甘肽和过氧化氢浓度。此外，实施了一种深度学习算法，利用二维拉曼光谱变换和多任务学习来提高测量速度和准确性。这种人工智能辅助的SERS方法可以在7分钟内识别胶质瘤患者的IDH1基因型。在31例胶质瘤患者队列中，该系统在准确区分IDH1基因型方面的受试者操作特征曲线下面积达到0.985。这项研究通过实现术中快速分子生物标志物识别，在完善手术决策和个性化术后治疗方面具有重大前景。
