Plasma sample cancer early screening method based on ensemble learning

Abstract

The invention discloses a plasma sample cancer early screening method based on ensemble learning, and belongs to the field of cancer early screening. The cancer early screening method comprises the following steps: 1, taking data obtained by performing characteristic value extraction on ctDNA mutation and methylation analysis data in plasma as a training set and a verification set, and then respectively importing the training set into a gradient boosting tree model and a classification model of a support vector machine; 2, fusing the gradient boosting tree model trained in the step 1 and the classification model of the support vector machine trained in the step 1 to obtain an integrated classification model; 3, importing the verification set in the step 1 into the integrated classification model in the step 3, and obtaining a classification result through a voting mechanism, namely, a cancer early screening result. The performance of the model is optimized under different training conditions, the adaptability to sample size, sample feature distribution and the like during model training is enhanced, the stability of the model is effectively improved, the reliability in practical application is ensured, and stable prediction precision is generated.

Description

technical field [0001] The invention belongs to the field of early cancer screening, in particular to a plasma sample early cancer screening method based on integrated learning. Background technique [0002] Early diagnosis of cancer is one of the most effective means to improve the survival rate of cancer patients. At present, early screening and clinical diagnosis of cancer mainly rely on CT, PET-CT, MRI and other imaging methods, as well as ultrasound, endoscopy, cytology detection, invasive tissue sample collection and pathological detection. The detection is mostly based on one or a combination of the above detection methods, such as CT and X-ray technology for high-risk groups of lung cancer. However, due to the small size of tumors in the early stage of cancer, CT identification requires the scanning area of ​​the diseased tissue to be larger than 1 cm, so CT, X-ray, and ultrasound examinations are difficult to identify early tumors. In addition, the accumulation of...

AI Technical Summary

Problems solved by technology

It can be seen that the current methods for diagnosing cancer have great limitations, and there are still defects such as inability to effectively detect early cancer, high cost, and poor tolerance of detection methods.

Images