Liver ultrasonic image identification method based on sparse expression

An ultrasound image and sparse representation technology, which is applied in the field of liver ultrasound image recognition based on sparse representation, and can solve the problems of complex and changeable space-occupying lesions.

CN105956620AInactive Publication Date: 2016-09-21SOUTH CHINA UNIV OF TECH
1 Cites 17 Cited by

Patent Information

Authority / Receiving Office
CN · China
Current Assignee / Owner
Publication Date
2016-09-21
Estimated Expiration
Not applicable · inactive patent

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
  • Figure 3
    Figure 3
Patent Text Reader

Abstract

The invention discloses a liver ultrasonic image identification method based on sparse expression. The method comprises the following steps of: (1) selecting an interested area from a liver ultrasonic image training sample with a space-occupying lesion area; (2) extracting a gray scale symbiosis matrix texture ratio characteristic, a fractal characteristic and a abrupt change rate characteristic of the interested area; (3) utilizing a dictionary expansion method based on sparse reconstruction to construct an expanded dictionary by the image characteristics obtained in the step (2); (4) utilizing the expanded dictionary obtained in the step (3) to construct a classifier based on sparse expression; and (5) inputting the image characteristics of a testing sample into the classifier for identification and judgment, and identifying out the liver ultrasonic image with the space-occupying lesion area. According to the invention, the classified identification accuracy is high, and each index accords with a clinical diagnosis range.
Need to check novelty before this filing date? Find Prior Art

Description

technical field

[0001] The invention relates to the technical field of ultrasonic image processing, in particular to a liver ultrasonic image recognition method based on sparse representation. Background technique

[0002] Liver cancer ranks sixth in cancer incidence and third in mortality worldwide. Early diagnosis of liver disease is conducive to the early detection and control of liver cancer, improving the survival rate of patients. Ultrasonography has the characteristics of no radiation, simple operation, repeatability, and low cost, so it is widely used in clinical diagnosis of liver diseases. Clinically, the diagnosis of ultrasound images of liver lesions relies on the naked eye observation of doctors to identify them. Not only is the workload huge, but the level of diagnosis depends to a certain extent on the experience of doctors. Therefore, it is of great significance to improve the overall level of ultrasound diagnosis by using medical image processing technolog...

Examples

Embodiment

[0068] As shown in Figure 1, a liver ultrasound image recognition method based on sparse representation includes the following steps:

[0069] (1) Select the region of interest from the liver ultrasound image training sample with the space-occupying lesion region, the region of interest includes the space-occupying lesion region R 1 and normal liver area R 2 The liver ultrasound image training samples include liver cyst image samples, hepatic hemangioma image samples, and liver cancer image samples, specifically including the following steps:

[0070] (1-1) Select the space-occupying lesion area R 1 : First, use the region-growing ultrasonic image automatic segmentation algorithm based on energy constraints to outline the edge of the lesion area, then take its circumscribed rectangle, and use the circumscribed rectangle area as the occupying lesion area R 1 ;

[0071] ROI (region of interest) refers to a region selected from the image, which is the focus of image analysis. ...