TY - JOUR
T1 - NHBS-Net: A feature fusion attention network for ultrasound neonatal hip bone segmentation
AU - Liu, Ruhan
AU - Liu, Mengyao
AU - Sheng, Bin
AU - Li, Huating
AU - Li, Ping
AU - Song, Haitao
AU - Zhang, Ping
AU - Jiang, Lixin
AU - Shen, Dinggang
N1 - Funding Information:
Manuscript received March 30, 2021; revised May 27, 2021 and May 31, 2021; accepted June 6, 2021. Date of publication June 9, 2021; date of current version November 30, 2021. This work was supported in part by the National Natural Science Foundation of China under Grant 61872241, Grant 61572316, and Grant 81771850; in part by the Science and Technology Commission of Shanghai Municipality under Grant 18410750700 and Grant 17411952600; in part by the Shanghai Jiao Tong University (SJTU) Medicine Engineering Interdisciplinary Research Fund under Grant YG2017MS19; and in part by The Hong Kong Polytechnic University under Grant P0030419, Grant P0030929, and Grant P0035358. (Ruhan Liu and Mengyao Liu contributed equally to this work.) (Corresponding authors: Bin Sheng; Lixin Jiang; Dinggang Shen.) Ruhan Liu and Bin Sheng are with the Department of Computer Science and Engineering, Shanghai Jiao Tong University, Shanghai 200240, China (e-mail: liurh996@sjtu.edu.cn; shengbin@sjtu.edu.cn).
Publisher Copyright:
© 1982-2012 IEEE.
PY - 2021/12
Y1 - 2021/12
N2 - Ultrasound is a widely used technology for diagnosing developmental dysplasia of the hip (DDH) because it does not use radiation. Due to its low cost and convenience, 2-D ultrasound is still the most common examination in DDH diagnosis. In clinical usage, the complexity of both ultrasound image standardization and measurement leads to a high error rate for sonographers. The automatic segmentation results of key structures in the hip joint can be used to develop a standard plane detection method that helps sonographers decrease the error rate. However, current automatic segmentation methods still face challenges in robustness and accuracy. Thus, we propose a neonatal hip bone segmentation network (NHBS-Net) for the first time for the segmentation of seven key structures. We design three improvements, an enhanced dual attention module, a two-class feature fusion module, and a coordinate convolution output head, to help segment different structures. Compared with current state-of-the-art networks, NHBS-Net gains outstanding performance accuracy and generalizability, as shown in the experiments. Additionally, image standardization is a common need in ultrasonography. The ability of segmentation-based standard plane detection is tested on a 50-image standard dataset. The experiments show that our method can help healthcare workers decrease their error rate from 6%-10% to 2%. In addition, the segmentation performance in another ultrasound dataset (fetal heart) demonstrates the ability of our network.
AB - Ultrasound is a widely used technology for diagnosing developmental dysplasia of the hip (DDH) because it does not use radiation. Due to its low cost and convenience, 2-D ultrasound is still the most common examination in DDH diagnosis. In clinical usage, the complexity of both ultrasound image standardization and measurement leads to a high error rate for sonographers. The automatic segmentation results of key structures in the hip joint can be used to develop a standard plane detection method that helps sonographers decrease the error rate. However, current automatic segmentation methods still face challenges in robustness and accuracy. Thus, we propose a neonatal hip bone segmentation network (NHBS-Net) for the first time for the segmentation of seven key structures. We design three improvements, an enhanced dual attention module, a two-class feature fusion module, and a coordinate convolution output head, to help segment different structures. Compared with current state-of-the-art networks, NHBS-Net gains outstanding performance accuracy and generalizability, as shown in the experiments. Additionally, image standardization is a common need in ultrasonography. The ability of segmentation-based standard plane detection is tested on a 50-image standard dataset. The experiments show that our method can help healthcare workers decrease their error rate from 6%-10% to 2%. In addition, the segmentation performance in another ultrasound dataset (fetal heart) demonstrates the ability of our network.
KW - medical image segmentation
KW - Neonatal hip bone segmentation
KW - self-attention mechanism
UR - http://www.scopus.com/inward/record.url?scp=85111054028&partnerID=8YFLogxK
U2 - 10.1109/TMI.2021.3087857
DO - 10.1109/TMI.2021.3087857
M3 - Journal article
C2 - 34106849
AN - SCOPUS:85111054028
SN - 0278-0062
VL - 40
SP - 3446
EP - 3458
JO - IEEE Transactions on Medical Imaging
JF - IEEE Transactions on Medical Imaging
IS - 12
ER -