Related Experiment Video
Updated: Jul 6, 2025
04:48
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
2.8K
DDCNN-F: double decker convolutional neural network 'F' feature fusion as a medical image classification framework
Nirmala Veeramani1, Premaladha Jayaraman2, Raghunathan Krishankumar3
1School of Computing, SASTRA Deemed to Be University, Thanjavur, India.
Scientific Reports
|January 5, 2024
Summary
A new Double Decker Convolutional Neural Network (DDCNN) feature fusion framework with an 'F' Flag feature improves melanoma classification. This approach enhances early detection of malignant melanoma, outperforming existing methods.
Area of Science:
- Dermatology
- Medical Imaging
- Artificial Intelligence
Background:
- Melanoma is a dangerous skin cancer requiring accurate early detection.
- Current classification methods face challenges with image analysis, particularly hairy lesions.
Purpose of the Study:
- To introduce a novel feature fusion framework for enhanced melanoma classification.
- To develop a new 'F' Flag feature for improved early detection of malignant melanoma.
Main Methods:
- Utilized a Double Decker Convolutional Neural Network (DDCNN) architecture.
- Incorporated a Convolutional Neural Network (CNN) for hairy image analysis using an intra-class variance score.
- Merged bottleneck features with ABCDE clinical indicators and the novel 'F' Flag feature for classification.
Main Results:
- Achieved high performance metrics: 98.4% specificity, 93.75% accuracy, 98.56% precision, and 0.98 AUC.
- The 'F' indicator improved classifier performance, increasing specificity by 7.34%.
Conclusions:
- The DDCNN 'F' Feature fusion framework offers a superior approach for melanoma diagnosis.
- This method demonstrates potential for accurate identification and diagnosis of fatal skin cancer.
