Method for diagnosing brain tumor and cerebral infarction

The invention discloses a method for diagnosing brain tumor and cerebral infarction. The method comprises the following steps: acquiring Raman spectrum data of brain tumor and cerebral infarction samples through a Raman spectrometer; preprocessing the obtained Raman spectrum data by using baseline correction, smoothing and normalization methods; building a convolutional neural network classification model, training the preprocessed Raman spectrum data, and adjusting model parameters to obtain an optimal convergence model; and inputting the Raman spectrum data of a to-be-diagnosed sample into the constructed convolutional neural network model, and outputting the category of a to-be-detected target. According to the invention, a judgment basis is provided for diagnosis of brain tumor and cerebral infarction by doctors, and life health of patients is further guaranteed. 本发明公开了一种诊断脑肿瘤和脑梗死的方法，包括通过拉曼光谱仪采集脑肿瘤和脑梗死样本的拉曼光谱数据；使用基线校正、平滑处理和归一化方法对获取的拉曼光谱数据进行预处理；搭建卷积神经网络分类模型，对预处理后的拉曼光谱数据进行训练，调节模型参数得到收敛最佳模型；将待诊断样本的拉曼光谱数据输入所搭