Evaluation of Deep Learning Methods for Bone Suppression from Dual Energy Chest Radiography
Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review
Standard
Evaluation of Deep Learning Methods for Bone Suppression from Dual Energy Chest Radiography. / Sirazitdinov, Ilyas; Kubrak, Konstantin; Kiselev, Semen; Tolkachev, Alexey; Kholiavchenko, Maksym; Ibragimov, Bulat.
Artificial Neural Networks and Machine Learning – ICANN 2020 - 29th International Conference on Artificial Neural Networks, Proceedings. ed. / Igor Farkaš; Paolo Masulli; Stefan Wermter. Springer VS, 2020. p. 247-257 (Lecture Notes in Computer Science, Vol. 12396 LNCS).Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - GEN
T1 - Evaluation of Deep Learning Methods for Bone Suppression from Dual Energy Chest Radiography
AU - Sirazitdinov, Ilyas
AU - Kubrak, Konstantin
AU - Kiselev, Semen
AU - Tolkachev, Alexey
AU - Kholiavchenko, Maksym
AU - Ibragimov, Bulat
N1 - Publisher Copyright: © 2020, Springer Nature Switzerland AG.
PY - 2020
Y1 - 2020
N2 - Bone suppression in chest x-rays is an important processing step that can often improve visual detection of lung pathologies hidden under ribs or clavicle shadows. Current diagnostic imaging protocol does not include hardware-based bone suppression, hence the need for a software-based solution. This paper evaluates various deep learning models adapted for bone suppression task, namely, we implemented several state-of-the-art deep learning architectures: convolution autoencoder, U-net, FPN, cGAN; augmented them with domain-specific denoising techniques, such as wavelet decomposition, with the aim to identify the optimal solution for chest x-ray analysis. Our results show that wavelet decomposition does not improve the rib suppression, “skip connections” modification outperforms baseline autoencoder approach with and without the usage of the wavelet decomposition, the residual models are trained faster than plain models and achieve higher validation scores.
AB - Bone suppression in chest x-rays is an important processing step that can often improve visual detection of lung pathologies hidden under ribs or clavicle shadows. Current diagnostic imaging protocol does not include hardware-based bone suppression, hence the need for a software-based solution. This paper evaluates various deep learning models adapted for bone suppression task, namely, we implemented several state-of-the-art deep learning architectures: convolution autoencoder, U-net, FPN, cGAN; augmented them with domain-specific denoising techniques, such as wavelet decomposition, with the aim to identify the optimal solution for chest x-ray analysis. Our results show that wavelet decomposition does not improve the rib suppression, “skip connections” modification outperforms baseline autoencoder approach with and without the usage of the wavelet decomposition, the residual models are trained faster than plain models and achieve higher validation scores.
KW - Bone shadow exclusion
KW - Bone suppression
KW - Chest x-ray
KW - Convolutional neural networks
KW - Deep learning
UR - http://www.scopus.com/inward/record.url?scp=85096520213&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-61609-0_20
DO - 10.1007/978-3-030-61609-0_20
M3 - Article in proceedings
AN - SCOPUS:85096520213
SN - 9783030616083
T3 - Lecture Notes in Computer Science
SP - 247
EP - 257
BT - Artificial Neural Networks and Machine Learning – ICANN 2020 - 29th International Conference on Artificial Neural Networks, Proceedings
A2 - Farkaš, Igor
A2 - Masulli, Paolo
A2 - Wermter, Stefan
PB - Springer VS
T2 - 29th International Conference on Artificial Neural Networks, ICANN 2020
Y2 - 15 September 2020 through 18 September 2020
ER -
ID: 271604995