TY - JOUR
T1 - S2FLNet
T2 - Hepatic steatosis detection network with body shape
AU - Wang, Qiyue
AU - Xue, Wu
AU - Zhang, Xiaoke
AU - Jin, Fang
AU - Hahn, James
N1 - Publisher Copyright:
© 2021 Elsevier Ltd
PY - 2022/1
Y1 - 2022/1
N2 - Fat accumulation in the liver cells can increase the risk of cardiac complications and cardiovascular disease mortality. Therefore, a way to quickly and accurately detect hepatic steatosis is critically important. However, current methods, e.g., liver biopsy, magnetic resonance imaging, and computerized tomography scan, are subject to high cost and/or medical complications. In this paper, we propose a deep neural network to estimate the degree of hepatic steatosis (low, mid, high) using only body shapes. The proposed network adopts dilated residual network blocks to extract refined features of input body shape maps by expanding the receptive field. Furthermore, to classify the degree of steatosis more accurately, we create a hybrid of the center loss and cross entropy loss to compact intra-class variations and separate inter-class differences. We performed extensive tests on the public medical dataset with various network parameters. Our experimental results show that the proposed network achieves a total accuracy of over 82% and offers an accurate and accessible assessment for hepatic steatosis.
AB - Fat accumulation in the liver cells can increase the risk of cardiac complications and cardiovascular disease mortality. Therefore, a way to quickly and accurately detect hepatic steatosis is critically important. However, current methods, e.g., liver biopsy, magnetic resonance imaging, and computerized tomography scan, are subject to high cost and/or medical complications. In this paper, we propose a deep neural network to estimate the degree of hepatic steatosis (low, mid, high) using only body shapes. The proposed network adopts dilated residual network blocks to extract refined features of input body shape maps by expanding the receptive field. Furthermore, to classify the degree of steatosis more accurately, we create a hybrid of the center loss and cross entropy loss to compact intra-class variations and separate inter-class differences. We performed extensive tests on the public medical dataset with various network parameters. Our experimental results show that the proposed network achieves a total accuracy of over 82% and offers an accurate and accessible assessment for hepatic steatosis.
KW - Center loss
KW - Dilated residual network
KW - Hepatic steatosis
UR - http://www.scopus.com/inward/record.url?scp=85120485730&partnerID=8YFLogxK
U2 - 10.1016/j.compbiomed.2021.105088
DO - 10.1016/j.compbiomed.2021.105088
M3 - Article
C2 - 34864582
AN - SCOPUS:85120485730
SN - 0010-4825
VL - 140
JO - Computers in Biology and Medicine
JF - Computers in Biology and Medicine
M1 - 105088
ER -