Lung adenocarcinoma subtype identification method and device based on image omics and deep learning

• Main Authors: ZHOU JING, FU XIAOTONG, JI YING
• Format: Patent
• Language: Chinese; English
• Published: 18.07.2023
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; PHYSICS

The invention discloses a lung adenocarcinoma subtype identification method and device based on radiomics and deep learning. The method comprises the following steps: acquiring a CT image with a nodule center marked in advance; for the CT image, image omics feature extraction and depth feature extraction are respectively carried out to obtain image omics features and depth features; the image omics features and the depth features are fused in the multi-head attention feature fusion step, and then the probability prediction result of each lung adenocarcinoma subtype is obtained through average pooling and an activation function. According to the method, the radiomics features and the depth features can be effectively integrated, and experimental results prove the effectiveness of the method in a pulmonary nodule classification task. 本发明公开了一种基于影像组学和深度学习的肺腺癌亚型识别方法和装置。该方法包括如下步骤：获得事先标注结节中心的CT图像；针对CT图像，分别进行影像组学特征提取和深度特征提取，获得影像组学特征和深度特征；影像组学特征和深度特征在多头注意力特征融合步骤中进行融合，再通过平均池化和激活函数得到各肺腺癌亚型的概率预测结果。本发明能够有效整合影像组学特征和深度特征，实验