Related Experiment Video
Updated: May 30, 2026
06:34
SCAnED - An Open-source Skin Segmentation Macro for Semi-automated Cell and Nuclei Detection in Epidermal and Dermal Skin Compartments
Published on: August 8, 2025
Automatic segmentation of dermoscopic images by iterative classification
Maciel Zortea1, Stein Olav Skrøvseth, Thomas R Schopf
1Department of Mathematics and Statistics, University of Tromsø, 9037 Tromsø, Norway.
International Journal of Biomedical Imaging
|August 4, 2011
Summary
This study introduces an automatic skin lesion segmentation method for dermoscopic images. The approach enhances accuracy, particularly for low-contrast lesions, improving computer-aided diagnosis.
Area of Science:
- Dermatology
- Medical Imaging
- Computer Vision
Background:
- Accurate skin lesion border detection is crucial for computer-aided diagnostic systems.
- Dermoscopic image analysis presents challenges, especially with low contrast between lesions and skin.
Purpose of the Study:
- To present a novel automatic approach for skin lesion segmentation in dermoscopic images.
- To improve the reliability and accuracy of lesion segmentation, particularly for challenging low-contrast cases.
Main Methods:
- An automatic rule derives seed regions for initial training samples based on image acquisition assumptions.
- Lesion segmentation is treated as a binary classification problem.
- An iterative hybrid classification strategy combines linear and quadratic classifiers to refine segmentation and training samples.
Main Results:
- The method automatically selects seed regions for reliable initial training.
- The iterative hybrid classification enhances segmentation accuracy.
- Improved performance is noted for challenging images with low contrast between skin and lesions.
Conclusions:
- The proposed automatic segmentation method offers a reliable approach for dermoscopic images.
- This technique is particularly beneficial for improving diagnostic accuracy in cases with subtle lesion borders.
- The approach enhances the robustness of computer-aided diagnostic systems for skin lesion analysis.
