Related Experiment Video
Updated: Sep 25, 2025
04:48
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
3.0K
A Novel Hybrid Deep Learning Approach for Skin Lesion Segmentation and Classification.
Puneet Thapar1, Manik Rakhra1, Gerardo Cazzato2
1Department of Computer Science and Engineering, Lovely Professional University, Punjab, India.
Journal of Healthcare Engineering
|April 28, 2022
Summary
This study introduces an AI approach using deep learning and swarm intelligence for accurate skin cancer diagnosis from dermoscopy images. The method significantly improves the ability to distinguish between benign and malignant skin lesions.
Area of Science:
- Dermatology
- Artificial Intelligence
- Medical Imaging
Background:
- Skin cancer detection relies on visual and dermoscopic analysis.
- Artificial intelligence, particularly deep learning, shows promise in analyzing skin images.
- Accurate discrimination between benign and malignant lesions is crucial for effective treatment.
Purpose of the Study:
- To develop a reliable AI-driven approach for skin cancer diagnosis using dermoscopy images.
- To enhance healthcare professionals' diagnostic capabilities in identifying malignant skin lesions.
- To improve the accuracy of differentiating benign from malignant skin lesions.
Main Methods:
- Utilized swarm intelligence (SI) algorithms for skin lesion region of interest (RoI) segmentation.
- Employed speeded-up robust features (SURF) for feature extraction from segmented RoIs.
- Applied Convolutional Neural Network (CNN) for classifying skin lesions into benign or malignant categories using ISIC-2017, ISIC-2018, and PH-2 datasets.
Main Results:
- Achieved an average classification accuracy of 98.42%.
- Reported an average precision of 97.73% and an MCC of 0.9704.
- Demonstrated superior performance across all evaluated metrics compared to previous methods.
Conclusions:
- The proposed AI-based segmentation and classification technique offers a reliable method for skin cancer diagnosis.
- This approach can significantly aid healthcare professionals in their diagnostic decisions.
- The study highlights the potential of integrating swarm intelligence and deep learning for improved dermatological diagnostics.
