Automatic segmentation of echocardiography based on a morphological reconstruction algorithm.

OBJECTIVE: To improve the precision of the traditional segmentation of echocardiogram, by suppressing the influence from inherent speckle noises in medical ultrasonic images. METHOD: An automatic segmentation method based on reconstructed morphology was proposed in this paper. First, the opening and closing operations by reconstruction were imposed to the ultrasonic image. Second, the top-hat operation was used to extract the bright and/or dark features and to find out the boundaries corresponding to these features, whereby implemented the automatic segmentation. RESULT: The segmented echocardiogram had less artificial boundaries resulted from speckle noise, and could accurately be extracted the artery and ventricle. CONCLUSION: The presented method can detect both dark and bright objects accurately, and the boundary has a fine continuity. In addition, the algorithm is also applicable to the extraction of sole bright/dark features, accordingly to reduce the complexity and time needed and to improve the accuracy.

[A medical ultrasonic image filtering method based on anisotropic diffusion].

OBJECTIVE: To remove the speckle noise in ultrasonic images by using anisotropic diffusion method. METHOD: Based on anisotropic diffusion, a partial differential equation, of which the initial data was the input images, was transformed into differential forms and solved with iterations. The speckle scale function of the equation was modified to make better use in filtering medical ultrasonic images. RESULT: By comparing the results with other three filters, the anisotropic diffusion method could smooth the speckles very well and the edge of the image was also clear. CONCLUSION: Anisotropic diffusion can remove the speckle noise effectively and has great potential in filtering medical ultrasonic images.

Quantitative analysis of sinoatrial node using Doppler tissue images.

A method has been developed to quantitatively analyze sinoatrial nodes (SAN) using Doppler tissue images (DTI). Doppler tissue images of SAN are acquired using an intracardiac catheter via the superior vena cava in an in vivo experiment. A sequence of DTI images of a SAN is obtained, and a complete cycle of the SAN excitation is observed. The tissue acceleration of the SAN is extracted and quantitatively analyzed. The estimated time-acceleration curve of the SAN exhibits remarkable similarity to the electrocardiogram curve. This is the first report on such finding. The experimental results show that the tissue movement of the SAN correlates with electrical cardiac activities and closely associates with the different phases of the cardiac cycle. This method has great potential in characterizing the local cardiac activities through the study of the conduct pathway.

[Research of myocardial velocity characteristics of paced canine heart based on ultrasonic Doppler tissue imaging].

OBJECTIVE: To quantitatively analyze the characteristics of regional myocardial velocities of paced heart over His bundle, and explore the myocardial velocity pattern. METHOD: 5 dogs were paced at 120 times/min over His bundle. We defined the His bundle adjacent region and the top region of interventricular septum (IVS) as our regions of interest (ROI). Images of Doppler tissue velocity in ROI and corresponding electrocardiograms were observed and recorded by a 2-D Color Doppler tissue imaging (DTI) system. Quantitative velocities of myocardial DTI images were mapped by a neural network from the image sequences. Finally the velocity characteristics of the myocardium around His bundle and IVS were analyzed. RESULT: We obtained myocardial time-velocity curves in the ROI, and some results based on those curves: In the His bundle adjacent region intervals of myocardial systole and diastole were (209.75 +/- 19.41) ms and (294.50 +/- 17.87) ms, respectively. Myocardial systole velocity (S), early diastole velocity (E) and late diastole velocity (A) were (5.43 +/- 1.53) cm/s, (3.95 +/- 1.18) cm/s and (1.72 +/- 0.77) cm/s, respectively, E/A was 2.06 +/- 0.95; In IVS region intervals of myocardial systole and diastole were (191.33 +/- 17.23) ms and (294.70 +/- 17.91) ms, respectively. S, E and A were (3.17 +/- 1.34) cm/s, (2.17 +/- 0.64) cm/s and (1.48 +/- 0.52) cm/s, respectively. E/A was 1.305 +/- 0.287. CONCLUSION: The primary characteristics of myocardial velocities paced over His bundle were: 1) Myocardial velocities in His bundle adjacent region were higher than those in IVS. 2) E/A was greater than one.

[Dynamic 3D reconstruction of Doppler flow ultrasound medical images].

OBJECTIVE: To implement dynamic 3D reconstruction of ultrasonic flow images by the technique of 3D reconstruction of ultrasonic medical images combined with Doppler flow imaging method. METHOD: Through analyzing the color-coded mode of Doppler flow images and utilizing the Color Bar in DFI provided by ultrasound system, the information of anatomical structure and blood flow velocity were separated from original DFI images. Then 3D reconstruction of blood velocity and its fusion display with 3D anatomical structure was implemented. RESULT: Clinical experiments showed that in vivo blood flow in heart cavity could be shown in 3D space simultaneously with anatomical structure of the heart, and the relationship between them was consistent with cardiology. The images rendered in such a way could show more medical information than that in traditional methods. CONCLUSION: The combination of 3D reconstruction of ultrasonic medical images and Doppler imaging technique could realize functional 3D reconstruction of ultrasonic medical images so as to give more medical information. It would have great potential application and present a new future for ultrasonic medical imaging.