Method for constructing pulmonary nodule computer-aided detection model

Abstract

The present invention discloses a method for constructing a pulmonary nodule computer-aided detection model. The method comprises the following steps: obtaining bronchial epithelial cells of a subjectand extracting RNA; constructing a double-stranded cDNA library according to the RNA and conducting sequencing; comparing a sequencing result with a reference genome and selecting a differential significant gene and a differential significant variation of the subject; using a vector formed by combining the differential significant gene and the differential significant variation, and benign and malignant pulmonary nodules as sample data; randomly dividing the sample data of the subject into a training set and a prediction set and carrying out multiple times of training on the training set andthe prediction set on the basis of a support vector machine model adopting a radial basis kernel function to obtain a model output value of the prediction set; and carrying out non-dimensionalizationon clinical data and CT image data of the subject in the prediction set, deducing a Roc fitting curve by combining a non-dimensionalized data comprehensive value and the model output value, calculating a lower area, and adjusting covariance and penalty factors to be the value corresponding to the maximum value of the lower area to obtain the pulmonary nodule computer-aided detection model.

Description

technical field [0001] The present invention relates to the field of computer technology. More specifically, it relates to a method for constructing a computer-aided detection model of small pulmonary nodules. Background technique [0002] Pulmonary nodules are defined as round or irregular lesions with a diameter of ≤2 cm in the lung, and imaging shows shadows with increased density, which can be single or multiple, with clear or unclear borders. Most diseases of the human lungs can lead to the formation of nodules, and pulmonary nodules mainly include benign lesions and malignant lesions. Among them, benign lesions mainly include hamartoma, hemangioma, inflammatory pseudotumor, and tuberculoma, while malignant lesions mainly refer to primary lung cancer diseases such as bronchioloalveolar carcinoma and lung adenocarcinoma, or malignant tumors in other parts of the body. transfer phenomenon. Relevant clinical studies have shown that 80% to 90% of nodules are benign accor...

AI Technical Summary

Problems solved by technology

Some patients need to obtain pathological tissue through surgery to differentiate between benign and malignant. The diagnosis rate is high, but the trauma is relatively large

Images