Automatic Recognition of Fetal Facial Standard Plane in Ultrasound Image via Fisher Vector.

Baiying Lei,Tianfu Wang,Dong Ni,Shengli Li,Ee-Leng Tan,Siping Chen,Liu Zhuo,Natasha M Maurits

doi:10.1371/journal.pone.0121838

Abstract

Acquisition of the standard plane is the prerequisite of biometric measurement and diagnosis during the ultrasound (US) examination. In this paper, a new algorithm is developed for the automatic recognition of the fetal facial standard planes (FFSPs) such as the axial, coronal, and sagittal planes. Specifically, densely sampled root scale invariant feature transform (RootSIFT) features are extracted and then encoded by Fisher vector (FV). The Fisher network with multi-layer design is also developed to extract spatial information to boost the classification performance. Finally, automatic recognition of the FFSPs is implemented by support vector machine (SVM) classifier based on the stochastic dual coordinate ascent (SDCA) algorithm. Experimental results using our dataset demonstrate that the proposed method achieves an accuracy of 93.27% and a mean average precision (mAP) of 99.19% in recognizing different FFSPs. Furthermore, the comparative analyses reveal the superiority of the proposed method based on FV over the traditional methods.

Highlights

Due to the relatively low cost, real-time imaging capability, and avoidance of radiation exposure [1,2,3,4], ultrasound (US) has been widely used for pregnancy diagnosis
A new algorithm is developed for the automatic recognition of the fetal facial standard planes (FFSPs) such as the axial, coronal, and sagittal planes
Automatic recognition of the FFSPs is implemented by support vector machine (SVM) classifier based on the stochastic dual coordinate ascent (SDCA) algorithm

Summary

Introduction

Due to the relatively low cost, real-time imaging capability, and avoidance of radiation exposure [1,2,3,4], ultrasound (US) has been widely used for pregnancy diagnosis. During the USbased diagnosis progress, the clinician first identifies the standard plane by checking the existence of main anatomical structures, and examines the plane for further diagnosis and interpretation of the fetal growth. The acquisition of standard plane requires substantial experience as well as good knowledge of the human anatomy. This task is extremely challenging for novices and time consuming even for experienced examiners. Accurate automatic recognition of standard plane is extremely useful in assisting both.

Objectives

Methods

Results