Similarity calculation system and method based on medical image features

Abstract

The invention provides a similarity calculation system and method based on medical image features. The system comprises a patient medical image library, a normal medical image library, a disease feature library, an image feature extraction module and a similarity calculation module. The patient medical image library is used for storing medical images of patients to be diagnosed. The normal medical image library is used for storing medical images of normal persons. The disease feature library is used for storing disease image features and relative diseases symptoms and disease history information inputted by a doctor. The image feature extraction module is used for receiving and extracting the medical images of the patients and the medical images of the normal persons to compare and find lesions areas, extracting image features of the lesions areas, and sending to the similarity calculation module. The similarity calculation module is used for comparing the received image features of the lesions areas with the image features in the disease feature library, and calculating similarity, obtaining a similarity calculating result, and sending to a display module to display. By the aid of the similarity calculation system, accuracy and efficiency of diagnosis are improved effectively, and working load of doctors is reduced greatly.

Description

technical field [0001] The present invention relates to the technical field of medical image diagnosis, in particular to a similarity calculation system and method based on medical image features. Background technique [0002] Using medical imaging to diagnose patients is a widely used diagnostic method. However, at present, doctors directly observe the disease through naked eyes. When manual diagnosis is performed, due to the large amount of image data and human eyes The image recognition ability is limited, misdiagnosis and missed diagnosis will inevitably occur, and various subjective mistakes will also appear from time to time, leaving a great hidden danger to the accuracy of patient treatment. Moreover, ct and pet examinations in hospitals often produce a large number of pictures, which makes the tasks of a limited number of medical imaging staff heavy and labor-intensive. Contents of the invention [0003] In view of the defects in the prior art, the object of the p...

AI Technical Summary

Problems solved by technology

[0002] Using medical imaging to diagnose patients is a widely used diagnostic method. However, at present, doctors directly observe the disease through naked eyes. When manual diagnosis is performed, due to the large amount of image data and human eyes The image recognition ability is limited, misdiagnosis and missed diagnosis will inevitably occur, and various subjective errors will also appear from time to time, leaving a great hidden danger to the accuracy of patient treatment

Moreover, ct and pet examinations in hospitals often produce a large number of pictures, which makes the task of a limited number of medical imaging staff heavy and labor-intensive

Images