A deep learning framework for automated early diagnosis and classification of skin cancer lesions in dermoscopy images
importance 8/10
partially solved
Cutaneous Melanoma Detection and Management skin cancer diagnosis stage: research
Problem
Manual diagnosis of skin cancer from dermoscopy images is challenging due to subtle visual differences between benign and malignant lesions.
Why it matters
Accurate, timely skin cancer diagnosis is crucial for successful treatment.
What the AI did
Skin-DeepNet segments lesions using Mask R-CNN and GrabCut, extracts features via HRNet with attention blocks, and classifies with a deep belief network plus decision fusion.
Results
Segmentation accuracy IOU up to 99.93%
Impact
The framework demonstrates automated lesion segmentation and classification on dermoscopy images.
Record
| Venue | Scientific Reports |
|---|---|
| Published | 2025 |
| Work type | article |
| Authors | Alaa S. Al‐Waisy, Shumoos Al-Fahdawi, Mohammed Khalaf, Mazin Abed Mohammed, Bourair Al-Attar, Mohammed Nasser Al-Andoli |
| Citations | 40 (OpenAlex, at import time) |
| References | 46 |
| Model or method | Skin-DeepNet (Mask R-CNN, HRNet, DBN) |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1038/s41598-025-15655-9 |
| Open access | gold |
| Keywords | deep learning, skin cancer, dermoscopy, segmentation, classification, diagnosis |
Importance rating
8
major — millions of lives, an industry or a field
Abstract calls skin cancer a significant global health concern with accurate, timely diagnosis crucial for successful treatment; affects millions worldwide.