We report an accurate, precise and sensitive method and system for quantitative fluorescence image-guided neurosurgery. With a low-noise, high-dynamic-range CMOS array, we perform rapid (integration times as low as 50 ms per wavelength) hyperspectral fluorescence and diffuse reflectance detection and apply a correction algorithm to compensate for the distorting effects of tissue absorption and scattering. Using this approach, we generated quantitative wide-field images of fluorescence in tissue-simulating phantoms for the fluorophore PpIX, having concentrations and optical absorption and scattering variations over clinically relevant ranges. The imaging system was tested in a rodent model of glioma, detecting quantitative levels down to 20 ng/ml. The resulting performance is a significant advance on existing wide-field quantitative imaging techniques, and provides performance comparable to a point-spectroscopy probe that has previously demonstrated significant potential for improved detection of malignant brain tumors during surgical resection.
展开▼
机译：我们报告了定量荧光图像引导的神经外科手术的准确，精确和敏感的方法和系统。使用低噪声，高动态范围的CMOS阵列，我们可以执行快速（每个波长的积分时间低至50毫秒）高光谱荧光和漫反射检测，并应用校正算法来补偿组织吸收和散射的扭曲效应。使用这种方法，我们在组织模拟体模中生成了荧光团PpIX的荧光定量广域图像，在临床相关范围内具有浓度，光吸收和散射变化。该成像系统在神经胶质瘤的啮齿动物模型中进行了测试，可检测低至20 ng / ml的定量水平。所产生的性能是现有广域定量成像技术的一项重大进步，并提供了与点光谱探针相当的性能，该点光谱探针先前已证明在外科手术切除过程中具有改善恶性脑肿瘤检测的巨大潜力。
展开▼
