Related Experiment Video
Updated: May 21, 2026
05:39
Dermoscopy Aids in the Diagnosis of Discoid Lupus Erythematosus
Published on: May 16, 2025
Lesion border detection in dermoscopy images using ensembles of thresholding methods
M Emre Celebi1, Quan Wen, Sae Hwang
1Department of Computer Science, Louisiana State University, Shreveport, LA, USA. ecelebi@lsus.edu
Summary
Automated analysis of dermoscopy images for skin lesion border detection is improved by using an ensemble of thresholding methods. This approach offers a robust and accurate solution for diagnosing melanoma and other pigmented skin lesions.
Area of Science:
- Dermatology
- Medical Imaging
- Computer Vision
Background:
- Dermoscopy is crucial for diagnosing skin lesions like melanoma.
- Automated analysis of dermoscopy images is vital due to interpretation subjectivity.
- Accurate lesion border detection is a critical first step in automated analysis.
Purpose of the Study:
- To develop an automated method for detecting lesion borders in dermoscopy images.
- To address the limitations of single thresholding methods in handling diverse dermoscopy images.
Main Methods:
- An ensemble of thresholding methods was employed for automated border detection.
- The method was applied to the blue channel of dermoscopy images.
Main Results:
- The proposed ensemble method demonstrated robustness across a challenging dataset.
- The method achieved high accuracy and speed compared to existing techniques.
Conclusions:
- Ensemble thresholding provides a robust, fast, and accurate solution for dermoscopy image analysis.
- This method enhances the automated diagnosis of melanoma and other pigmented skin lesions.
