Related Experiment Video
Updated: May 27, 2025
09:37
Combining Reflectance Confocal Microscopy with Optical Coherence Tomography for Noninvasive Diagnosis of Skin Cancers via Image Acquisition
Published on: August 18, 2022
2.2K
A fuzzy rank-based deep ensemble methodology for multi-class skin cancer classification
Arindam Halder1, Anogh Dalal1, Sanghita Gharami1
1Department of Information Technology, Jadavpur University, Jadavpur University Salt Lake Campus, Plot No. 8, Salt Lake Bypass, LB Block, Sector III, Salt Lake City, Kolkata, 700106, West Bengal, India.
Scientific Reports
|February 20, 2025
Summary
This study developed an effective skin cancer detection model using deep learning and fuzzy logic on the HAM10000 dataset. The approach achieved 95.14% accuracy, demonstrating potential for early skin cancer identification.
Area of Science:
- Dermatology and Medical Imaging
- Artificial Intelligence in Healthcare
Background:
- Skin cancer is a significant global health concern, identified by the World Health Organisation (WHO) as a leading cause of mortality.
- Early detection of skin cancer is crucial for effective treatment and improved patient outcomes.
Purpose of the Study:
- To develop and evaluate an accurate skin cancer classification model using deep learning techniques.
- To assess the model's performance on benchmark datasets for clinical applicability.
Main Methods:
- Utilized the HAM10000 dataset, involving image augmentation for dataset balancing, preprocessing, and splitting into training, testing, and validation sets.
- Trained individual deep learning models (Xception, InceptionResNetV2, MobileNetV2) and combined their predictions using fuzzy logic for final classification.
- Evaluated model performance using metrics such as classification accuracy and confusion matrix.
Main Results:
- Achieved an impressive classification accuracy of 95.14% on the HAM10000 dataset.
- Successfully validated the model on the DermaMNIST dataset, surpassing the benchmark accuracy (78.25% vs. 76.8%).
Conclusions:
- The developed ensemble model demonstrates high efficiency and accuracy in classifying skin cancer lesions.
- The model shows significant potential for integration into clinical applications for early skin cancer diagnosis.
