autoPET IV challenge: Incorporating organ supervision and human guidance for lesion segmentation in PET/CT

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Huang, Junwei, Hao, Yingqi, Luo, Yitong, Wang, Ziyu, Liu, Mingxuan, Chen, Yifei, Wang, Yuanhan, Xiang, Lei, Tian, Qiyuan
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866918134421651456
author Huang, Junwei
Hao, Yingqi
Luo, Yitong
Wang, Ziyu
Liu, Mingxuan
Chen, Yifei
Wang, Yuanhan
Xiang, Lei
Tian, Qiyuan
author_facet Huang, Junwei
Hao, Yingqi
Luo, Yitong
Wang, Ziyu
Liu, Mingxuan
Chen, Yifei
Wang, Yuanhan
Xiang, Lei
Tian, Qiyuan
contents Lesion Segmentation in PET/CT scans is an essential part of modern oncological workflows. To address the challenges of time-intensive manual annotation and high inter-observer variability, the autoPET challenge series seeks to advance automated segmentation methods in complex multi-tracer and multi-center settings. Building on this foundation, autoPET IV introduces a human-in-the-loop scenario to efficiently utilize interactive human guidance in segmentation tasks. In this work, we incorporated tracer classification, organ supervision and simulated clicks guidance into the nnUNet Residual Encoder framework, forming an integrated pipeline that demonstrates robust performance in a fully automated (zero-guidance) context and efficiently leverages iterative interactions to progressively enhance segmentation accuracy.
format Preprint
id arxiv_https___arxiv_org_abs_2509_02402
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle autoPET IV challenge: Incorporating organ supervision and human guidance for lesion segmentation in PET/CT
Huang, Junwei
Hao, Yingqi
Luo, Yitong
Wang, Ziyu
Liu, Mingxuan
Chen, Yifei
Wang, Yuanhan
Xiang, Lei
Tian, Qiyuan
Image and Video Processing
Lesion Segmentation in PET/CT scans is an essential part of modern oncological workflows. To address the challenges of time-intensive manual annotation and high inter-observer variability, the autoPET challenge series seeks to advance automated segmentation methods in complex multi-tracer and multi-center settings. Building on this foundation, autoPET IV introduces a human-in-the-loop scenario to efficiently utilize interactive human guidance in segmentation tasks. In this work, we incorporated tracer classification, organ supervision and simulated clicks guidance into the nnUNet Residual Encoder framework, forming an integrated pipeline that demonstrates robust performance in a fully automated (zero-guidance) context and efficiently leverages iterative interactions to progressively enhance segmentation accuracy.
title autoPET IV challenge: Incorporating organ supervision and human guidance for lesion segmentation in PET/CT
topic Image and Video Processing
url https://arxiv.org/abs/2509.02402