Automatic Fetal Ultrasound Standard Plane Detection Using Knowledge Transferred Recurrent Neural Networks
Accurate acquisition of fetal ultrasound (US) standard planes is one of the most crucial steps in obstetric diagnosis. The conventional way of standard plane acquisition requires a thorough knowledge of fetal anatomy and intensive manual labors. Hence, au
- PDF / 2,412,796 Bytes
- 8 Pages / 439.363 x 666.131 pts Page_size
- 17 Downloads / 204 Views
Dept. of Computer Science and Engineering, The Chinese University of Hong Kong 2 School of Medicine, Shenzhen University, China 3 Shenzhen Maternal and Child Healthcare Hospital of Nanfang Medical University
Abstract. Accurate acquisition of fetal ultrasound (US) standard planes is one of the most crucial steps in obstetric diagnosis. The conventional way of standard plane acquisition requires a thorough knowledge of fetal anatomy and intensive manual labors. Hence, automatic approaches are highly demanded in clinical practice. However, automatic detection of standard planes containing key anatomical structures from US videos remains a challenging problem due to the high intra-class variations of standard planes. Unlike previous studies that developed specific methods for different anatomical standard planes respectively, we present a general framework to detect standard planes from US videos automatically. Instead of utilizing hand-crafted visual features, our framework explores spatio-temporal feature learning with a novel knowledge transferred recurrent neural network (T-RNN), which incorporates a deep hierarchical visual feature extractor and a temporal sequence learning model. In order to extract visual features effectively, we propose a joint learning framework with knowledge transfer across multi-tasks to address the insufficiency issue of limited training data. Extensive experiments on different US standard planes with hundreds of videos corroborate that our method can achieve promising results, which outperform state-of-the-art methods.
1
Introduction
Obstetric ultrasound (US) examination generally involves the procedures of image scanning, standard plane selection, biometric measurement and diagnosis. Accurate acquisition of US standard planes, e.g., fetal abdominal standard plane (FASP), fetal face axial standard plane (FFASP) and fetal four-chamber view standard plane (FFVSP) of heart, is one crucial step for the subsequent biometric measurement and obstetric diagnosis. Clinically, US standard plane is manually acquired by searching the view with concurrent presence of key anatomical structures (KASs) in the regions of interest (ROI) [1]. Fig. 1 illustrates the KASs for
Corresponding author.
c Springer International Publishing Switzerland 2015 N. Navab et al. (Eds.): MICCAI 2015, Part I, LNCS 9349, pp. 507–514, 2015. DOI: 10.1007/978-3-319-24553-9_62
508
H. Chen et al.
Nose bone Lens
LV
SP UV
RV
Eyes
LA
SB RA
DAO
Fig. 1. Left: FFASP containing nose bone, eyes and lens; middle: FASP containing stomach bubble (SB), umbilical vein (UV) and spine (SP); right: FFVSP containing left atrium (LA), right atrium (RA), left ventricle (LV), right ventricle (RV) and descending aorta (DAO) (green rectangles denote the ROIs).
FFASP, FASP and FFVSP, respectively. The manual acquisition of standard planes heavily relies on clinical experience and is also very laborious. Hence, automatic detection methods are highly demanded to boost the examination efficiency [2]. However, this computerized detection ta
Data Loading...
