TY - GEN
T1 - Multi-modal Pathological Pre-training via Masked Autoencoders for Breast Cancer Diagnosis
AU - Lu, Mengkang
AU - Wang, Tianyi
AU - Xia, Yong
N1 - Publisher Copyright:
© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023.
PY - 2023
Y1 - 2023
N2 - Breast cancer (BC) is one of the most common cancers identified globally among women, which has become the leading cause of death. Multi-modal pathological images contain different information for BC diagnosis. Hematoxylin and eosin (H &E) staining images could reveal a considerable amount of microscopic anatomy. Immunohistochemical (IHC) staining images provide the evaluation of the expression of various biomarkers, such as the human epidermal growth factor receptor (HER2) hybridization. In this paper, we propose a multi-modal pre-training model via pathological images for BC diagnosis. The proposed pre-training model contains three modules: (1) the modal-fusion encoder, (2) the mixed attention, and (3) the modal-specific decoders. The pre-trained model could be performed on multiple relevant tasks (IHC Reconstruction and IHC classification). The experiments on two datasets (HEROHE Challenge and BCI Challenge) show state-of-the-art results.
AB - Breast cancer (BC) is one of the most common cancers identified globally among women, which has become the leading cause of death. Multi-modal pathological images contain different information for BC diagnosis. Hematoxylin and eosin (H &E) staining images could reveal a considerable amount of microscopic anatomy. Immunohistochemical (IHC) staining images provide the evaluation of the expression of various biomarkers, such as the human epidermal growth factor receptor (HER2) hybridization. In this paper, we propose a multi-modal pre-training model via pathological images for BC diagnosis. The proposed pre-training model contains three modules: (1) the modal-fusion encoder, (2) the mixed attention, and (3) the modal-specific decoders. The pre-trained model could be performed on multiple relevant tasks (IHC Reconstruction and IHC classification). The experiments on two datasets (HEROHE Challenge and BCI Challenge) show state-of-the-art results.
KW - Breast cancer
KW - Hematoxylin and eosin staining
KW - Immunohistochemical staining
KW - Multi-modal pre-training
UR - http://www.scopus.com/inward/record.url?scp=85174692545&partnerID=8YFLogxK
U2 - 10.1007/978-3-031-43987-2_44
DO - 10.1007/978-3-031-43987-2_44
M3 - 会议稿件
AN - SCOPUS:85174692545
SN - 9783031439865
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 457
EP - 466
BT - Medical Image Computing and Computer Assisted Intervention – MICCAI 2023 - 26th International Conference, Proceedings
A2 - Greenspan, Hayit
A2 - Greenspan, Hayit
A2 - Madabhushi, Anant
A2 - Mousavi, Parvin
A2 - Salcudean, Septimiu
A2 - Duncan, James
A2 - Syeda-Mahmood, Tanveer
A2 - Taylor, Russell
PB - Springer Science and Business Media Deutschland GmbH
T2 - 26th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2023
Y2 - 8 October 2023 through 12 October 2023
ER -
