Octree Boundary Transfiner: Efficient Transformers for Tumor Segmentation Refinement

Anthony Wang, Ti Bai, Dan Nguyen, Steve Jiang

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Scopus citations

Abstract

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.

Original languageEnglish (US)
Title of host publicationHead and Neck Tumor Segmentation and Outcome Prediction - 3rd Challenge, HECKTOR 2022, Held in Conjunction with MICCAI 2022, Proceedings
EditorsVincent Andrearczyk, Valentin Oreiller, Adrien Depeursinge, Mathieu Hatt
PublisherSpringer Science and Business Media Deutschland GmbH
Pages54-60
Number of pages7
ISBN (Print)9783031274190
DOIs
StatePublished - 2023
Event3rd 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 - Singapore, Singapore
Duration: Sep 22 2022Sep 22 2022

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume13626 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference3rd 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
Country/TerritorySingapore
CitySingapore
Period9/22/229/22/22

Keywords

  • Octree
  • Transformer
  • U-Net

ASJC Scopus subject areas

  • Theoretical Computer Science
  • General Computer Science

Fingerprint

Dive into the research topics of 'Octree Boundary Transfiner: Efficient Transformers for Tumor Segmentation Refinement'. Together they form a unique fingerprint.

Cite this