AutoPET Challenge: Combining nn-Unet with Swin UNETR Augmented by Maximum Intensity Projection Classifier

Abstract

Tumor volume and changes in tumor characteristics over time are important biomarkers for cancer therapy. In this context, FDG-PET/CT scans are routinely used for staging and re-staging of cancer, as the radiolabeled fluorodeoxyglucose is taken up in regions of high metabolism. Unfortunately, these regions with high metabolism are not specific to tumors and can also represent physiological uptake by normal functioning organs, inflammation, or infection, making detailed and reliable tumor segmentation in these scans a demanding task. This gap in research is addressed by the AutoPET challenge, which provides a public data set with FDG-PET/CT scans from 900 patients to encourage further improvement in this field. Our contribution to this challenge is an ensemble of two state-of-the-art segmentation models, the nn-Unet and the Swin UNETR, augmented by a maximum intensity projection classifier that acts like a gating mechanism. If it predicts the existence of lesions, both segmentations are combined by a late fusion approach. Our solution achieves a Dice score of 72.12% on patients diagnosed with lung cancer, melanoma, and lymphoma in our cross-validation. Code: https://github.com/heiligerl/autopet_submission

Publication
arXiv
Lars Heiliger
Lars Heiliger
PhD Student
Jana Fragemann
Jana Fragemann
PhD Student
Jacob Murray
Jacob Murray
MD Candidate
Rainer Stiefelhagen
Rainer Stiefelhagen
Director Computer Vision for Human-Computer Interaction Lab
Jens Kleesiek
Jens Kleesiek
Professor of Translational Image-guided Oncology