Improved graph cut model with features of superpixels and neighborhood patches for myocardium segmentation from ultrasound image.

Xiangfen Song,Qing Wang,Qianjin Feng,Yinong Wang

doi:10.3934/mbe.2019053

Xiangfen Song, Qing Wang + Show 2 more

Open Access

https://doi.org/10.3934/mbe.2019053

Copy DOI

Abstract

Ultrasound (US) imaging has the technical advantages for the functional evaluation of myocardium compared with other imaging modalities. However, it is a challenge of extracting the myocardial tissues from the background due to low quality of US imaging. To better extract the myocardial tissues, this study proposes a semi-supervised segmentation method of fast Superpixels and Neighborhood Patches based Continuous Min-Cut (fSP-CMC). The US image is represented by a graph, which is constructed depending on the features of superpixels and neighborhood patches. A novel similarity measure is defined to capture and enhance the features correlation using Pearson correlation coefficient and Pearson distance. Interactive labels provided by user play a subsidiary role in the semi-supervised segmentation. The continuous graph cut model is solved via a fast minimization algorithm based on augmented Lagrangian and operator splitting. Additionally, Non-Uniform Rational B-Spline (NURBS) curve fitting is used as post-processing to solve the low resolution problem caused by the graph-based method. 200 B-mode US images of left ventricle of the rats were collected in this study. The myocardial tissues were segmented using the proposed fSP-CMC method compared with the method of fast Neighborhood Patches based Continuous Min-Cut (fP-CMC). The results show that the fSP-CMC segmented the myocardial tissues with a higher agreement with the ground truth (GT) provided by medical experts. The mean absolute distance (MAD) and Hausdorff distance (HD) were significantly lower than those values of fP-CMC (p < 0.05), while the Dice was significantly higher (p < 0.05). In conclusion, the proposed fSP-CMC method accurately and effectively segments the myocardiumn in US images. This method has potentials to be a reliable segmentation method and useful for the functional evaluation of myocardium in the future study.

Highlights

Myocardial infarction (MI) is a severe cardio-vascular disease that threatens human health, leading to the subsequent death of cardiomyocytes and vascular cells in the vicinity site of the infarction
The myocardial segmentation results of the proposed fSP-CMC method were in high agreement with the ground truth (GT) results in the parasternal long-axis (PLAX)-viewed and parasternal short-axis (PSAX)-viewed images of left ventricle of rats (Figure 6)
It was found that the receiver operating characteristic (ROC) curves obtained from the fSP-CMC segmented images were located closer to the upper left corner indicating a higher agreement in the segmentation results with the GT

Summary

Introduction

Myocardial infarction (MI) is a severe cardio-vascular disease that threatens human health, leading to the subsequent death of cardiomyocytes and vascular cells in the vicinity site of the infarction. In the typical B-mode ultrasound (US) images, the structure of the heart and the morphology of the myocardium can be checked. The left ventricle, an important pumping chamber of the heart, can be scanned in the parasternal long-axis (PLAX) and parasternal short-axis (PSAX) views showing the cavity (the hypoechoic area) and the myocardial tissue (isoechoic area surrounding the cavity). Some inherent drawbacks of US imaging, such as low contrast, speckle noise, signal dropout, acoustic shadow, cause the myocardial tissue indistinguishable from the background. It is challenging to investigate accurate and effective segmentation algorithms of myocardium ultrasound (MUS) images. Segmentation of target tissue from other tissues or background is an essential phase of ultrasound computer-aided diagnosis [1]. Myocardial segmentation and assessment are the keys in the morphology and function study of myocardium [2]

Methods

Results

Conclusion