A Localization-to-Segmentation Framework for Automatic Tumor Segmentation in Whole-Body PET/CT Images

TL;DR

This paper introduces L2SNet, a two-stage localization-to-segmentation framework for accurate tumor segmentation in PET/CT images, addressing challenges posed by small tumor size and high-uptake normal tissues.

Contribution

The paper presents a novel localization-to-segmentation approach with an adaptive threshold scheme, improving tumor segmentation accuracy in PET/CT images.

Findings

Achieved top 7 ranking in MICCAI 2023 challenge

Demonstrated improved segmentation performance

Validated effectiveness of the adaptive threshold scheme

Abstract

Fluorodeoxyglucose (FDG) positron emission tomography (PET) combined with computed tomography (CT) is considered the primary solution for detecting some cancers, such as lung cancer and melanoma. Automatic segmentation of tumors in PET/CT images can help reduce doctors' workload, thereby improving diagnostic quality. However, precise tumor segmentation is challenging due to the small size of many tumors and the similarity of high-uptake normal areas to the tumor regions. To address these issues, this paper proposes a localization-to-segmentation framework (L2SNet) for precise tumor segmentation. L2SNet first localizes the possible lesions in the lesion localization phase and then uses the location cues to shape the segmentation results in the lesion segmentation phase. To further improve the segmentation performance of L2SNet, we design an adaptive threshold scheme that takes the segmentation results of the two phases into consideration. The experiments with the MICCAI 2023 Automated Lesion Segmentation in Whole-Body FDG-PET/CT challenge dataset show that our method achieved a competitive result and was ranked in the top 7 methods on the preliminary test set. Our work is available at: https://github.com/MedCAI/L2SNet.