TY - GEN
T1 - Octree Boundary Transfiner
T2 - 3rd 3D Head and Neck Tumor Segmentation in PET/CT Challenge, HECKTOR 2022, held in Conjunction with the 25th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2022
AU - Wang, Anthony
AU - Bai, Ti
AU - Nguyen, Dan
AU - Jiang, Steve
N1 - Publisher Copyright:
© 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.
PY - 2023
Y1 - 2023
N2 - In this paper, we create a fully autonomous system that segments primary head and neck tumors as well as lymph node tumors given only FDG-PET and CT scans without contrast enhancers. Given only these two modalities, the typical Dice score for the state-of-the-art (SOTA) models lies below 0.8, below what it would be when including other modalities due to the low resolution of PET scans and noisy non-enhanced CT images. Thus, we seek to improve tumor segmentation accuracy while working with the limitation of only having these two modalities. We introduce the Transfiner, a novel octree-based refinement system to harness the fidelity of transformers while keeping computation and memory costs low for fast inferencing. The observation behind our method is that segmentation errors almost always occur at the edges of a mask for predictions from a well-trained model. The Transfiner utilizes base network feature maps in addition to the raw modalities as input and selects regions of interest from these. These are then processed with a transformer network and decoded with a CNN. We evaluated our framework with Dice Similarity Coefficient (DSC) 0.76426 for the first task of the Head and Neck Tumor Segmentation Challenge (HECKTOR) and ranked 6th.
AB - In this paper, we create a fully autonomous system that segments primary head and neck tumors as well as lymph node tumors given only FDG-PET and CT scans without contrast enhancers. Given only these two modalities, the typical Dice score for the state-of-the-art (SOTA) models lies below 0.8, below what it would be when including other modalities due to the low resolution of PET scans and noisy non-enhanced CT images. Thus, we seek to improve tumor segmentation accuracy while working with the limitation of only having these two modalities. We introduce the Transfiner, a novel octree-based refinement system to harness the fidelity of transformers while keeping computation and memory costs low for fast inferencing. The observation behind our method is that segmentation errors almost always occur at the edges of a mask for predictions from a well-trained model. The Transfiner utilizes base network feature maps in addition to the raw modalities as input and selects regions of interest from these. These are then processed with a transformer network and decoded with a CNN. We evaluated our framework with Dice Similarity Coefficient (DSC) 0.76426 for the first task of the Head and Neck Tumor Segmentation Challenge (HECKTOR) and ranked 6th.
KW - Octree
KW - Transformer
KW - U-Net
UR - http://www.scopus.com/inward/record.url?scp=85151057475&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85151057475&partnerID=8YFLogxK
U2 - 10.1007/978-3-031-27420-6_5
DO - 10.1007/978-3-031-27420-6_5
M3 - Conference contribution
AN - SCOPUS:85151057475
SN - 9783031274190
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 54
EP - 60
BT - Head and Neck Tumor Segmentation and Outcome Prediction - 3rd Challenge, HECKTOR 2022, Held in Conjunction with MICCAI 2022, Proceedings
A2 - Andrearczyk, Vincent
A2 - Oreiller, Valentin
A2 - Depeursinge, Adrien
A2 - Hatt, Mathieu
PB - Springer Science and Business Media Deutschland GmbH
Y2 - 22 September 2022 through 22 September 2022
ER -