Researchers used big data to predict lymph node metastasis in lung cancer

reporter learned from the Suzhou Medical Institute of the Chinese Academy of Sciences, the institute's researchers in cooperation with Shanghai Long March Hospital, using big data analysis, recently developed a more accurate prediction of lung cancer lymph node metastasis evaluation model. Trials in more than 300 cancer patients showed that the evaluation model was even more predictive than clinicians.
, researchers combined big data technology with medical imaging to extract information from imagery for deep analysis. The team selected non-small cell lung cancer, which accounted for more than 80% of the number of lung cancer patients, as the study subjects, and analyzed CT image data of 366 cancer patients.
Xin, a researcher at Suzhou Medical Institute who was involved in the study, said that most of the traditional tumor imaging evaluation focused only on tumor areas, and the biggest innovation of the new model was the inclusion of image features around tumors. Tumor surrounding environment secretes a large number of growth factors and cytokines, induces hypoxia and angiogenesic, plays an important role in tumor occurrence, development and metastasis, and can further improve the accuracy and scientificity of prediction by combining tumor and tumor peripheral imaging characteristics.
366 lung cancer patients showed that the imaging characteristics of the area around the tumor could be used to predict lymph node metastasis in early lung cancer, the team said. The prediction accuracy reached 82.5% by the new model based on the imaging characteristics of tumor and tumor peripheral area. Combining big data assessment with doctor's judgment to build a "human-machine combination" model, the prediction accuracy is further increased to more than 85%.
study can be used not only to predict lymph node metastasis in non-small cell lung cancer, but also to a variety of other types of cancer. The high accuracy of predictions obtained by combining big data with doctor's diagnosis also suggests that 'human-machine integration' is likely to be a new trend in clinical development in the future. Gao Xin said. (Source: Xinhua News Agency,
results have recently been published in the field of radiology in the world's top journal, European Radiology.
relevant paper information: