State Key Laboratory of Tribology, Tsinghua University, Beijing 100084, China.
Department of Pathology, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing 100021, China.
Spectrochim Acta A Mol Biomol Spectrosc. 2022 Jan 15;265:120400. doi: 10.1016/j.saa.2021.120400. Epub 2021 Sep 14.
Intraoperative detection of the marginal tissues is the last and most important step to complete the resection of adenocarcinoma and squamous cell carcinoma. However, the current intraoperative diagnosis is time-consuming and requires numerous steps including staining. In this paper, we present the use of Raman spectroscopy with deep learning to achieve accurate diagnosis with stain-free process. To make the spectrum more suitable for deep learning, we utilize an unusual way of thinking which regards Raman spectral signal as a sequence and then converts it into two-dimensional Raman spectrogram by short-time Fourier transform as input. The normal-adenocarcinoma deep learning model and normal-squamous carcinoma deep learning model both achieve more than 96% accuracy, 95% sensitivity and 98% specificity when test, which higher than the conventional principal components analysis-linear discriminant analysis method with normal-adenocarcinoma model (0.896 accuracy, 0.867 sensitivity, 0.926 specificity) and normal-squamous carcinoma model (0.821 accuracy, 0.776 sensitivity, 1.000 specificity). The high performance of deep learning models provides a reliable way for intraoperative detection of marginal tissue, and is expected to reduce the detection time and save human lives.
术中检测边缘组织是完成腺癌和鳞状细胞癌切除的最后也是最重要的步骤。然而，目前的术中诊断既耗时又需要多个步骤，包括染色。在本文中，我们提出使用拉曼光谱与深度学习相结合，实现无染色过程的准确诊断。为了使光谱更适合深度学习，我们采用了一种不同寻常的思维方式，即将拉曼光谱信号视为一个序列，然后通过短时傅里叶变换将其转换为二维拉曼光谱图作为输入。在测试中，正常-腺癌深度学习模型和正常-鳞状细胞癌深度学习模型的准确率均超过 96%，灵敏度均超过 95%，特异性均超过 98%，高于传统的主成分分析-线性判别分析方法的正常-腺癌模型（准确率 0.896，灵敏度 0.867，特异性 0.926）和正常-鳞状细胞癌模型（准确率 0.821，灵敏度 0.776，特异性 1.000）。深度学习模型的高性能为边缘组织的术中检测提供了一种可靠的方法，有望减少检测时间，拯救生命。
