Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

CO2Wounds-V2: Extended Chronic Wounds Dataset From Leprosy Patients

Karen Sanchez, Carlos Hinojosa, Olinto Mieles, Chen Zhao, Bernard Ghanem, Henry Arguello

TL;DR

CO2Wounds-V2 delivers an extended, publicly available dataset of $764$ RGB wound images from $96$ leprosy patients, with semantic segmentation annotations in COCO format and binary masks to support automated wound analysis. The dataset, captured via smartphone in a developing-country clinical setting, is split into $485/122/157$ images for training, validation, and testing, respectively, with the testing set remaining unlabeled. Baseline experiments compare FPN, U-Net, and DeepLabV3+ architectures across multiple encoders, using $100$-epoch training with Adam, demonstrating performance in the $mIoU$ range of the high 60s to around 70, and highlighting the highest $mIoU$ and F1 scores for the U-Net baselines. The work provides a practical benchmark and underscores the potential of smartphone-based wound imaging for remote monitoring and algorithm development, while addressing ethical considerations and data licensing to enable widespread use in research and clinical contexts.

Abstract

Chronic wounds pose an ongoing health concern globally, largely due to the prevalence of conditions such as diabetes and leprosy's disease. The standard method of monitoring these wounds involves visual inspection by healthcare professionals, a practice that could present challenges for patients in remote areas with inadequate transportation and healthcare infrastructure. This has led to the development of algorithms designed for the analysis and follow-up of wound images, which perform image-processing tasks such as classification, detection, and segmentation. However, the effectiveness of these algorithms heavily depends on the availability of comprehensive and varied wound image data, which is usually scarce. This paper introduces the CO2Wounds-V2 dataset, an extended collection of RGB wound images from leprosy patients with their corresponding semantic segmentation annotations, aiming to enhance the development and testing of image-processing algorithms in the medical field.

CO2Wounds-V2: Extended Chronic Wounds Dataset From Leprosy Patients

TL;DR

CO2Wounds-V2 delivers an extended, publicly available dataset of RGB wound images from leprosy patients, with semantic segmentation annotations in COCO format and binary masks to support automated wound analysis. The dataset, captured via smartphone in a developing-country clinical setting, is split into images for training, validation, and testing, respectively, with the testing set remaining unlabeled. Baseline experiments compare FPN, U-Net, and DeepLabV3+ architectures across multiple encoders, using -epoch training with Adam, demonstrating performance in the range of the high 60s to around 70, and highlighting the highest and F1 scores for the U-Net baselines. The work provides a practical benchmark and underscores the potential of smartphone-based wound imaging for remote monitoring and algorithm development, while addressing ethical considerations and data licensing to enable widespread use in research and clinical contexts.

Abstract

Chronic wounds pose an ongoing health concern globally, largely due to the prevalence of conditions such as diabetes and leprosy's disease. The standard method of monitoring these wounds involves visual inspection by healthcare professionals, a practice that could present challenges for patients in remote areas with inadequate transportation and healthcare infrastructure. This has led to the development of algorithms designed for the analysis and follow-up of wound images, which perform image-processing tasks such as classification, detection, and segmentation. However, the effectiveness of these algorithms heavily depends on the availability of comprehensive and varied wound image data, which is usually scarce. This paper introduces the CO2Wounds-V2 dataset, an extended collection of RGB wound images from leprosy patients with their corresponding semantic segmentation annotations, aiming to enhance the development and testing of image-processing algorithms in the medical field.
Paper Structure (21 sections, 2 figures, 1 table)

This paper contains 21 sections, 2 figures, 1 table.

Table of Contents

  1. Introduction
  2. CO2Wounds-V2 Dataset
  3. Dataset description
  4. Materials
  5. Data collection
  6. Data curation
  7. Annotation
  8. Ethical aspects
  9. Autonomy
  10. Beneficence
  11. Non-maleficence
  12. Justice
  13. Respect
  14. License
  15. Evaluation
  16. ...and 6 more sections

Figures (2)

  • Figure 1: Random image samples from the proposed CO2Wounds-V2 dataset with their annotations. Rectangulars in each image: detection bounding boxes; shaded area in each image: segmentation masks.
  • Figure 2: Random samples from the validation set of our CO2Wounds-V2 dataset (Column 1), ground truth segmentation masks (Column 2), and segmentation results using U-Net architecture (Column 3).