Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Robust and Annotation-Free Wound Segmentation on Noisy Real-World Pressure Ulcer Images: Towards Automated DESIGN-R\textsuperscript{\textregistered} Assessment

Yun-Cheng Tsai

TL;DR

DESIGN-R wound assessment requires accurate segmentation, but foot-focused models struggle on non-foot wounds. The authors propose a detector–segmentation cascade using a YOLOv11n ROI detector trained on 500 bounding boxes combined with a fixed pre-trained FUSegNet, enabling zero fine-tuning across wound sites. Evaluated on 526 real-world images from foot, sacrum, and trochanter, the pipeline achieves 99.0% end-to-end success and substantial improvements in segmentation quality, with practical DESIGN-R mapping and plans to release detector artifacts. This annotation-efficient approach facilitates scalable wound monitoring in diverse clinical settings, reducing labeling burden and enabling deployment on low-resource devices.

Abstract

Purpose: Accurate wound segmentation is essential for automated DESIGN-R scoring. However, existing models such as FUSegNet, which are trained primarily on foot ulcer datasets, often fail to generalize to wounds on other body sites. Methods: We propose an annotation-efficient pipeline that combines a lightweight YOLOv11n-based detector with the pre-trained FUSegNet segmentation model. Instead of relying on pixel-level annotations or retraining for new anatomical regions, our method achieves robust performance using only 500 manually labeled bounding boxes. This zero fine-tuning approach effectively bridges the domain gap and enables direct deployment across diverse wound types. This is an advance not previously demonstrated in the wound segmentation literature. Results: Evaluated on three real-world test sets spanning foot, sacral, and trochanter wounds, our YOLO plus FUSegNet pipeline improved mean IoU by 23 percentage points over vanilla FUSegNet and increased end-to-end DESIGN-R size estimation accuracy from 71 percent to 94 percent (see Table 3 for details). Conclusion: Our pipeline generalizes effectively across body sites without task-specific fine-tuning, demonstrating that minimal supervision, with 500 annotated ROIs, is sufficient for scalable, annotation-light wound segmentation. This capability paves the way for real-world DESIGN-R automation, reducing reliance on pixel-wise labeling, streamlining documentation workflows, and supporting objective and consistent wound scoring in clinical practice. We will publicly release the trained detector weights and configuration to promote reproducibility and facilitate downstream deployment.

Robust and Annotation-Free Wound Segmentation on Noisy Real-World Pressure Ulcer Images: Towards Automated DESIGN-R\textsuperscript{\textregistered} Assessment

TL;DR

DESIGN-R wound assessment requires accurate segmentation, but foot-focused models struggle on non-foot wounds. The authors propose a detector–segmentation cascade using a YOLOv11n ROI detector trained on 500 bounding boxes combined with a fixed pre-trained FUSegNet, enabling zero fine-tuning across wound sites. Evaluated on 526 real-world images from foot, sacrum, and trochanter, the pipeline achieves 99.0% end-to-end success and substantial improvements in segmentation quality, with practical DESIGN-R mapping and plans to release detector artifacts. This annotation-efficient approach facilitates scalable wound monitoring in diverse clinical settings, reducing labeling burden and enabling deployment on low-resource devices.

Abstract

Purpose: Accurate wound segmentation is essential for automated DESIGN-R scoring. However, existing models such as FUSegNet, which are trained primarily on foot ulcer datasets, often fail to generalize to wounds on other body sites. Methods: We propose an annotation-efficient pipeline that combines a lightweight YOLOv11n-based detector with the pre-trained FUSegNet segmentation model. Instead of relying on pixel-level annotations or retraining for new anatomical regions, our method achieves robust performance using only 500 manually labeled bounding boxes. This zero fine-tuning approach effectively bridges the domain gap and enables direct deployment across diverse wound types. This is an advance not previously demonstrated in the wound segmentation literature. Results: Evaluated on three real-world test sets spanning foot, sacral, and trochanter wounds, our YOLO plus FUSegNet pipeline improved mean IoU by 23 percentage points over vanilla FUSegNet and increased end-to-end DESIGN-R size estimation accuracy from 71 percent to 94 percent (see Table 3 for details). Conclusion: Our pipeline generalizes effectively across body sites without task-specific fine-tuning, demonstrating that minimal supervision, with 500 annotated ROIs, is sufficient for scalable, annotation-light wound segmentation. This capability paves the way for real-world DESIGN-R automation, reducing reliance on pixel-wise labeling, streamlining documentation workflows, and supporting objective and consistent wound scoring in clinical practice. We will publicly release the trained detector weights and configuration to promote reproducibility and facilitate downstream deployment.

Paper Structure

This paper contains 24 sections, 3 figures, 8 tables.

Table of Contents

  1. Introduction
  2. Key Contributions
  3. Proposed Approach
  4. Problem Definition
  5. Segmentation as a Prerequisite for DESIGN-R Assessment
  6. Clinical Dataset and Evaluation Strategy
  7. Our Proposed Strategy: Combining Detection and Segmentation
  8. Proposed Method
  9. Overview of the Proposed Pipeline
  10. YOLOv11n Detector Training
  11. Evaluation on Clinical Dataset
  12. Results
  13. Dataset and Experiment Setup
  14. Pipeline Success Rate
  15. Segmentation Outcome and Visual Validation
  16. ...and 9 more sections

Figures (3)

  • Figure 1: Proposed wound segmentation pipeline: ROI detection using YOLOv11n followed by FUSegNet segmentation. The workflow enables the application of FUSegNet to non-foot clinical wound images.
  • Figure 2: Precision-Recall and F1-score curves comparing YOLOv11n and YOLOv11s. While YOLOv11s yields higher peak scores, YOLOv11n offers more stable detection across thresholds and is ultimately selected for deployment.
  • Figure 3: Representative qualitative segmentation results from the proposed YOLOv11n + FUSegNet pipeline on diverse wound types. Each row shows the input ROI, predicted segmentation mask, and overlay. Despite being trained only on foot ulcers, the pipeline generalizes effectively to sacrum, trochanter, and limb wounds.