In Physics in medicine and biology
The incidence of esophageal squamous cell carcinoma (ESCC) in China is high and the prognosis is poor. In order to evaluate the ESCC, we investigated computerized quantitative analysis on diagnostic computed tomography (CT) and digital histopathological slices. A retrospective study with IRB approval was conducted to assess the prognosis of ESCCs in 153 patients who underwent esophagectomy. Each of them has a 3D CT image and a pathological tissue slide after hematoxylin-eosin staining. Then, we performed quantitative analysis on digital histological slices and diagnostic CT volumes. We designed a set of quantitative features based on machine learning for pathological images. These features describe the patterns of different types of cells in pathological images and have good prognostic performance. At the same time, we also compared multiple machine learning models and adopted a five-fold cross-validation method to establish a more robust survival model. This paper combines CT images and pathological images to extract quantitative features from them for survival analysis, and achieves better prognostic results than using only clinical features and CT images.
Wang Jinlong, Wu Leilei, Zhang Yunzhe, Ma Guowei, Lu Yao
CT radiomics, Esophageal Squamous Cell Carcinoma(ESCC), feature extraction, histopathological image, pattern recognition, prognosis