A novel skin cancer detection model using modified finch deep CNN classifier model

Area of Science:

• Dermatology
• Artificial Intelligence
• Medical Imaging

Background:

• Skin cancer is a significant health concern, with rising mortality rates necessitating early detection.
• Current skin cancer detection methods face challenges in accuracy and computational efficiency.
• Ultraviolet radiation exposure is a primary cause of abnormal skin cell growth leading to skin cancer.

Purpose of the Study:

• To develop a novel and efficient method for skin cancer detection.
• To improve the accuracy and reduce the computational time of existing skin cancer detection techniques.
• To introduce a modified Falcon Finch deep Convolutional Neural Network (CNN) classifier for enhanced disease identification.

Main Methods:

• A modified Falcon Finch deep CNN classifier was developed for skin cancer detection.
• The Falcon Finch optimization algorithm was integrated for efficient parameter tuning of the deep CNN.
• The classifier was evaluated for its ability to analyze relevant skin cancer information and minimize errors.

Main Results:

• The modified Falcon Finch deep CNN achieved high accuracy (93.59% k-fold, 96.52% training), sensitivity (92.14% k-fold, 96.69% training), and specificity (95.22% k-fold, 96.54% training).
• The proposed method demonstrated enhanced robustness and faster convergence compared to existing approaches.
• The classifier effectively analyzed skin cancer data, minimizing diagnostic errors.

Conclusions:

• The modified Falcon Finch deep CNN offers a highly effective solution for skin cancer detection.
• This novel approach significantly improves diagnostic accuracy and efficiency, addressing limitations of current methods.
• The integration of Falcon Finch optimization enhances the performance and speed of deep CNNs in medical diagnostics.