TY - GEN
T1 - Collaborative learning of cross-channel clinical attention for radiotherapy-related esophageal fistula prediction from ct
AU - Cui, Hui
AU - Xu, Yiyue
AU - Li, Wanlong
AU - Wang, Linlin
AU - Duh, Henry
N1 - Publisher Copyright:
© Springer Nature Switzerland AG 2020.
PY - 2020
Y1 - 2020
N2 - Early prognosis of the radiotherapy-related esophageal fistula is of great significance in making personalized stratification and optimal treatment plans for esophageal cancer (EC) patients. The effective fusion of diagnostic consideration guided multi-level radiographic visual descriptors is a challenging task. We propose an end-to-end clinical knowledge enhanced multi-level cross-channel feature extraction and aggregation model. Firstly, clinical attention is represented by contextual CT, segmented tumor and anatomical surroundings from nine views of planes. Then for each view, a Cross-Channel-Atten Network is proposed with CNN blocks for multi-level feature extraction, cross-channel convolution module for multi-domain clinical knowledge embedding at the same feature level, and attentional mechanism for the final adaptive fusion of multi-level cross-domain radiographic features. The experimental results and ablation study on 558 EC patients showed that our model outperformed the other methods in comparison with or without multi-view, multi-domain knowledge, and multi-level attentional features. Visual analysis of attention maps shows that the network learns to focus on tumor and organs of interests, including esophagus, trachea, and mediastinal connective tissues.
AB - Early prognosis of the radiotherapy-related esophageal fistula is of great significance in making personalized stratification and optimal treatment plans for esophageal cancer (EC) patients. The effective fusion of diagnostic consideration guided multi-level radiographic visual descriptors is a challenging task. We propose an end-to-end clinical knowledge enhanced multi-level cross-channel feature extraction and aggregation model. Firstly, clinical attention is represented by contextual CT, segmented tumor and anatomical surroundings from nine views of planes. Then for each view, a Cross-Channel-Atten Network is proposed with CNN blocks for multi-level feature extraction, cross-channel convolution module for multi-domain clinical knowledge embedding at the same feature level, and attentional mechanism for the final adaptive fusion of multi-level cross-domain radiographic features. The experimental results and ablation study on 558 EC patients showed that our model outperformed the other methods in comparison with or without multi-view, multi-domain knowledge, and multi-level attentional features. Visual analysis of attention maps shows that the network learns to focus on tumor and organs of interests, including esophagus, trachea, and mediastinal connective tissues.
KW - Cross channel attention
KW - CT
KW - Esophageal fistula prediction
UR - http://www.scopus.com/inward/record.url?scp=85093121984&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-59710-8_21
DO - 10.1007/978-3-030-59710-8_21
M3 - Conference article published in proceeding or book
AN - SCOPUS:85093121984
SN - 9783030597092
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 212
EP - 220
BT - Medical Image Computing and Computer Assisted Intervention – MICCAI 2020 - 23rd International Conference, Proceedings
A2 - Martel, Anne L.
A2 - Abolmaesumi, Purang
A2 - Stoyanov, Danail
A2 - Mateus, Diana
A2 - Zuluaga, Maria A.
A2 - Zhou, S. Kevin
A2 - Racoceanu, Daniel
A2 - Joskowicz, Leo
PB - Springer Science and Business Media Deutschland GmbH
T2 - 23rd International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2020
Y2 - 4 October 2020 through 8 October 2020
ER -
