RETRACTED ARTICLE: DenseNet-II: an improved deep convolutional neural network for melanoma cancer detection

Area of Science:

• Dermatology and Medical Imaging
• Artificial Intelligence in Healthcare
• Computational Pathology

Background:

• Melanoma is a deadly cancer caused by uncontrolled melanocyte growth.
• Accurate melanoma detection relies on visual data like dermatological scans.
• Existing models often classify lesions into only benign or malignant categories.

Purpose of the Study:

• To propose a deep learning CNN framework for improved melanoma detection accuracy.
• To customize neural network architecture, activation functions, and input dimensions.
• To enhance the classification of seven distinct skin lesion types.

Main Methods:

• Extensive use of neural networks, specifically deep learning CNNs.
• Customization of network layers, activation functions, and input array dimensions.
• Utilized the HAM10000 dataset containing seven lesion types for training and validation.

Main Results:

• Achieved improved accuracy in melanoma detection through customized CNN models.
• Demonstrated effective classification across seven diverse skin lesion categories.
• Validated the efficacy of deep learning approaches for complex dermatological diagnosis.

Conclusions:

• The proposed deep learning CNN framework significantly enhances melanoma detection.
• Customizing network parameters is crucial for improving diagnostic precision.
• This approach offers a more diversified and accurate method for identifying various skin lesions.