Pulmonary tuberculosis classification diagnosis model training method based on artificial intelligence

Abstract

The invention discloses a pulmonary tuberculosis classification diagnosis model training method based on artificial intelligence. The method specifically comprises the steps of S1, collecting chest radiograph data of a patient; S2, dividing the collected chest radiograph data into a normal chest radiograph and a non-normal chest radiograph through a medical imaging expert; S3, inputting the data of the normal position chest radiograph and the non-normal position chest radiograph into a residual convolutional neural network for training, and stopping training when the training converges or reaches a set training round number to obtain a chest radiograph classification model; S4, further labeling the normal position chest radiograph judged in the step S2 by a medical imaging expert, and dividing the pulmonary tuberculosis chest radiograph into nine classes; and S4, inputting the anterior chest radiograph and nine-classification labels labeled by imaging experts into the convolutional detection neural network for training. According to the invention, whether the patient is tuberculosis or not can be diagnosed through imaging assistance based on the DR chest radiograph, and the chest radiograph of the patient with tuberculosis can be further distinguished.

Description

technical field [0001] The invention relates to the technical field of medical imaging aided diagnosis, in particular to an artificial intelligence-based training method for a pulmonary tuberculosis classification and diagnosis model. Background technique [0002] Tuberculosis is a major infectious disease that is under the control of our country. The epidemic situation in my country is severe, and the number of patients ranks second in the world. There are about 830,000 new tuberculosis patients every year. Chest X-ray is one of the effective methods for auxiliary diagnosis and screening of pulmonary tuberculosis. It is based on the manual interpretation of chest X-rays by imaging doctors. In some areas in the Midwest, there is a lack of well-trained radiologists, and X-ray chest X-rays cannot be used well. Application and interpretation; therefore, it is a very meaningful and effective work to use artificial intelligence diagnostic technology to assist in the diagnosis of ...

AI Technical Summary

Problems solved by technology

Therefore, merely diagnosing whether a patient has tuberculosis is increasingly not in line with clinical needs

Images