Medical image recognizing method

Abstract

The invention relates to a medical image recognition method, which aims at providing the method which can more accurately recognize the type of a new medical image. The method comprises the construction of a classification library of association rules, the update thereof and the medical image recognition step, the construction of the classification library of the association rules and the update step thereof comprise the following steps: data of medical image samples are prepared and carried out the pre-treatment; a density clustering-based medical image segmentation method is adopted to respectively recognize local region to be analyzed in each sample medical image; the characteristics of the local region in each sample medical image are respectively extracted to construct a medical image sample database T, the characteristics comprise: mean, variance, inclination, kurtosis, energy, entropy and clustering characteristics; the characteristic values are carried out the discretization; a frequent itemset in the medical image sample database is excavated; the classification library of the association rules is constructed according to the frequent itemset; and the frequent itemset is regularly or irregularly updated, meanwhile the classification library of the association rules is updated.

Description

technical field [0001] The invention belongs to the application field of computer analysis technology of medical images, and in particular relates to a method for medical image recognition. Background technique [0002] Since the 1970s, with the emergence and rapid development of medical imaging technologies such as computed tomography (CT), magnetic resonance imaging (MRI), and ultrasound US, a large number of medical images for clinical diagnosis and analysis have been generated and stored in hospitals. In recent years, with the rapid development of computers and related technologies and the maturity of graphics and image technology, people can arbitrarily zoom in, zoom out, rotate, contrast, adjust, segment, register, and three-dimensionally reconstruct medical images. Medical images can be observed from multiple directions, levels, and angles, thereby assisting doctors in key analysis of lesion bodies and other areas of interest, improving the accuracy of clinical diagno...

AI Technical Summary

Problems solved by technology

However, most of the current image diagnosis methods are still based on visually observing the lesion area in the image, and relying on the doctor's personal clinical experience and subjective judgment to make a diagnosis.

This method has the following deficiencies: ① The utilization rate of information is not high. Since medical images generally have a high resolution (such as CT images with a gray level of up to 4096), those image information and image features that cannot be distinguished by the human eye cannot be obtained. Full application will affect the early judgment of the disease; ②Doctors have personal subjectivity, and the diagnosis results largely depend on the doctor's personal clinical experience. For those doctors with less clinical experience, it is necessary to make the correct diagnosis. Diagnosis is very difficult

These methods more or less have the following problems: ①It is limited to mining the features of the whole image, while the local area features of human body images in clinical diagnosis have more clinical significance

The vast majority of feature extraction is based on the entire image, or simply divide the image into several regular parts, and extract the features of each part separately. These features cannot really represent the real information in the image, and the recognition effect The quality depends largely on the extracted features; ②The method for medical image classification still stays in the direct use of traditional data mining algorithms such as association rules. Medical image data is complex and high-dimensional, which needs to be studied A classification algorithm suitable for its characteristics; ③The recognition of medical images is limited to individual specific tissues and organs, such as breasts, brains, etc., and the research results obtained do not have general adaptability to the knowledge mining application of other organs; ④Methods The features taken are only some basic features such as color, texture, shape, etc., and the characteristics of the medical image itself have not been fully considered; ⑤The classification method takes a long time to train, has low accuracy, and is not practical

Images