CT (computed tomography) tuberculosis testing artificial intelligence diagnosis and treatment system

Abstract

The invention discloses a CT (computed tomography) tuberculosis testing artificial intelligence diagnosis and treatment system which comprises a feature sample library, a neuron network learning module, a deep convolutional neural network learning module, an output module and a computer terminal, and a large number of samples capable of being learned by the neuron network learning module and the deep convolutional neural network learning module are arranged in the feature sample library. The system is novel in design, the neuron network learning module and the deep convolutional neural networklearning module can deep learn marked images in the feature sample library by the aid of a pulmonary nodule artificial intelligence diagnosis and treatment system to obtain accurate pulmonary noduleclassification strategy, pulmonary nodules are detected, the sizes, the shapes and benign and malignant properties of the nodules are judged, a structured report is automatically made for reference ofdoctors, so that intelligent film reading and auxiliary diagnosis and treatment functions based on pulmonary nodule images are provided for the doctors, and diagnosis and treatment suggestions are offered for the doctors.

Description

technical field [0001] The invention relates to the field of medical technology, in particular to an artificial intelligence diagnosis and treatment system for CT pulmonary tuberculosis detection. Background technique [0002] Tuberculosis is a chronic infectious disease caused by Mycobacterium tuberculosis, which can invade many organs, with pulmonary tuberculosis infection being the most common. The excretor is an important source of infection. After the human body is infected with Mycobacterium tuberculosis, the disease may not necessarily occur. When the resistance is reduced or the cell-mediated allergic reaction is increased, it may cause clinical disease. Infiltrative pulmonary tuberculosis, the X-ray is usually cloudy or small sheet-like infiltrating shadows, blurred edges (exudative) or nodular, cord-like (proliferative) lesions, large consolidation or spherical lesions (caseous-visible cavities) ) or calcification, chronic fibrous cavitary pulmonary tuberculosis,...

AI Technical Summary

Problems solved by technology

Furthermore, this scheme has high recognition efficiency and high recognition accuracy, effectively avoids the phenomenon of missed detection and unrecognized, and effectively solves the problems existing in the prior art. Pulmonary nodules are classified to distinguish different quality CT images and various types of nodules, and no multi-channel, heterogeneous three-dimensional convolution fusion algorithm is used to improve the detection sensitivity and accuracy of different scales and different shapes of pulmonary nodules

Images