Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Lesion Segmentation in FDG-PET/CT Using Swin Transformer U-Net 3D: A Robust Deep Learning Framework

Shovini Guha, Dwaipayan Nandi

TL;DR

The paper tackles automated lesion segmentation in FDG-PET/CT, addressing the limitations of manual delineation and CNN-based methods in capturing long-range context. It introduces SwinUNet3D, a dual-channel transformer-based U-Net that fuses PET functional activity with CT anatomy using patch embedding and Swin Transformer blocks within a 3D encoder-decoder. The approach achieves Dice 0.88 and IoU 0.78 on AutoPET III, outperforming a 3D U-Net baseline, while delivering faster inference and improved handling of small or irregular lesions. This work demonstrates the practical potential of transformer-based architectures for robust, efficient PET/CT segmentation and sets the stage for multi-tracer and multi-center validation to support clinical deployment.

Abstract

Accurate and automated lesion segmentation in Positron Emission Tomography / Computed Tomography (PET/CT) imaging is essential for cancer diagnosis and therapy planning. This paper presents a Swin Transformer UNet 3D (SwinUNet3D) framework for lesion segmentation in Fluorodeoxyglucose Positron Emission Tomography / Computed Tomography (FDG-PET/CT) scans. By combining shifted window self-attention with U-Net style skip connections, the model captures both global context and fine anatomical detail. We evaluate SwinUNet3D on the AutoPET III FDG dataset and compare it against a baseline 3D U-Net. Results show that SwinUNet3D achieves a Dice score of 0.88 and IoU of 0.78, surpassing 3D U-Net (Dice 0.48, IoU 0.32) while also delivering faster inference times. Qualitative analysis demonstrates improved detection of small and irregular lesions, reduced false positives, and more accurate PET/CT fusion. While the framework is currently limited to FDG scans and trained under modest GPU resources, it establishes a strong foundation for future multi-tracer, multi-center evaluations and benchmarking against other transformer-based architectures. Overall, SwinUNet3D represents an efficient and robust approach to PET/CT lesion segmentation, advancing the integration of transformer-based models into oncology imaging workflows.

Lesion Segmentation in FDG-PET/CT Using Swin Transformer U-Net 3D: A Robust Deep Learning Framework

TL;DR

The paper tackles automated lesion segmentation in FDG-PET/CT, addressing the limitations of manual delineation and CNN-based methods in capturing long-range context. It introduces SwinUNet3D, a dual-channel transformer-based U-Net that fuses PET functional activity with CT anatomy using patch embedding and Swin Transformer blocks within a 3D encoder-decoder. The approach achieves Dice 0.88 and IoU 0.78 on AutoPET III, outperforming a 3D U-Net baseline, while delivering faster inference and improved handling of small or irregular lesions. This work demonstrates the practical potential of transformer-based architectures for robust, efficient PET/CT segmentation and sets the stage for multi-tracer and multi-center validation to support clinical deployment.

Abstract

Accurate and automated lesion segmentation in Positron Emission Tomography / Computed Tomography (PET/CT) imaging is essential for cancer diagnosis and therapy planning. This paper presents a Swin Transformer UNet 3D (SwinUNet3D) framework for lesion segmentation in Fluorodeoxyglucose Positron Emission Tomography / Computed Tomography (FDG-PET/CT) scans. By combining shifted window self-attention with U-Net style skip connections, the model captures both global context and fine anatomical detail. We evaluate SwinUNet3D on the AutoPET III FDG dataset and compare it against a baseline 3D U-Net. Results show that SwinUNet3D achieves a Dice score of 0.88 and IoU of 0.78, surpassing 3D U-Net (Dice 0.48, IoU 0.32) while also delivering faster inference times. Qualitative analysis demonstrates improved detection of small and irregular lesions, reduced false positives, and more accurate PET/CT fusion. While the framework is currently limited to FDG scans and trained under modest GPU resources, it establishes a strong foundation for future multi-tracer, multi-center evaluations and benchmarking against other transformer-based architectures. Overall, SwinUNet3D represents an efficient and robust approach to PET/CT lesion segmentation, advancing the integration of transformer-based models into oncology imaging workflows.
Paper Structure (25 sections, 3 equations, 2 figures, 3 tables)

This paper contains 25 sections, 3 equations, 2 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Methodology
  3. Data Acquisition and Preprocessing
  4. Dataset Description
  5. Preprocessing Steps
  6. Input Initialization
  7. Normalization
  8. Padding
  9. Stacking
  10. Patching
  11. Model Design and Architecture
  12. Input Format
  13. Architecture Overview
  14. Patch Embedding
  15. Encoder Block
  16. ...and 10 more sections

Figures (2)

  • Figure 1: Preprocessing workflow applied to PET/CT data. Raw inputs undergo intensity normalization, zero-padding, and patching before being fed into the SwinUNet3D network. This ensures consistent input dimensions and efficient utilization of 3D context.
  • Figure 2: Overview of the proposed SwinUNet3D architecture. The model adopts a U-Net-like encoder–decoder designb15 with hierarchical Swin Transformer blocksb13, patch embedding, bottleneck, and skip connections (see labels in the figure). This structure enables both local feature extraction and global context modeling for accurate 3D lesion segmentation.