Method and device for retrieving digital medical image

Abstract

The invention discloses a method and a device for retrieving a digital medical image. The method comprises the following steps: acquiring a searching instance; carrying out primary retrieval in a medical database according to the searching instance to acquire a plurality of primary retrieval images; detecting angle differences between the searching instance and each primary retrieval image by mutual information; aligning the searching instance with each primary retrieval image respectively according to the angle differences and centroid distance between the searching instance and the primary retrieval images; calculating distance between the searching instance and each aligned primary retrieval image; and selecting first N primary retrieval images with minimum distance value as a retrievedtarget image. The application of the method and the device can greatly improve the response speed of retrieval while maintaining excellent retrieval precision, and is more applicable to clinical actual environment.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a method and device for retrieving digital medical images. Background technique [0002] With the widespread application of digital medical imaging equipment such as CT, Magnetic Resonance Imaging (MRI, Magnetic Resonance Imaging), and ultrasound in clinical work, medical institutions generate a large amount of medical image data every day. Studies have shown that the amount of image data generated by the radiology department of a large hospital has reached TB level every year. The traditional keyword-based retrieval method faces the problem of inaccurate image data retrieval when retrieving medical data containing images. As a necessary supplement, content-based medical image retrieval (CBMIR, Content Based Medical Image Retrieval) important means of data retrieval. It plays an important auxiliary role in clinical diagnosis, medical teaching and medical research. [...

AI Technical Summary

Problems solved by technology

Although the rigid registration is used as a refined retrieval method, although the retrieval accuracy is high, the response speed is very slow, which cannot meet the actual needs and is difficult to be applied in the actual clinical environment.

Images