New coronal pneumonia intelligent diagnosis system based on deep learning

Abstract

The invention relates to an intelligent diagnosis system for new coronal pneumonia based on deep learning, and belongs to the field of medical image processing. The system comprises a control unit, an intelligent detection and diagnosis unit, a storage unit and a three-dimensional display unit, the control unit is used for inputting and modifying system data; the intelligent detection and diagnosis unit is used for predicting new coronal pneumonia cases; the intelligent detection and diagnosis unit comprises a data preprocessing module, a focus area detection module, a false positive removal module and a case prediction module. The deep learning network diagnosis system combining three parts of detection, false positive removal and prediction is adopted to output the diagnosis result, the problems that the focus is too small and is not easy to detect, the information of a single local focus is too little, the misdiagnosis rate is too high and the like are solved, the problem that the number of medical samples is too small is solved, the diagnosis efficiency is greatly improved, and the diagnosis accuracy is improved. Therefore, treatment efficiency of a patient can be improved, and clinical data can be accumulated.

Description

technical field [0001] The invention belongs to the field of medical image processing, and relates to an intelligent diagnosis system for new coronary pneumonia based on deep learning. Background technique [0002] The global novel coronavirus pneumonia (COVID-19) infectious disease has quickly attracted the attention of medical experts due to its high infectivity and high pathogenicity rate, and accurate diagnosis of patients is helpful for disease control and recovery. At present, there are two main methods for detecting new coronary pneumonia, nucleic acid detection and CT detection. However, due to the shortage of reagents and relatively high false negative and misdiagnosis rates in nucleic acid detection, CT detection is usually used as the main diagnostic basis. Usually, the CT of a case contains 200-500 tomographic images, and the traditional method of reading images for doctors is often time-consuming and laborious, and some lesions are small in size and difficult to...

AI Technical Summary

Problems solved by technology

At present, there are two main methods for detecting new coronary pneumonia: nucleic acid detection and CT detection. However, due to the shortage of reagents and the relatively high rate of false negative and misdiagnosis in nucleic acid detection, CT detection is usually used as the main dia...

Images