Sequential Convolutional Neural Network Techniques to Improve Detection of Lesions in Dogs

Background: Animal health and well-being depend on the early identification and treatment of diseases, especially skin conditions. Machine learning (ML) techniques, such as convolutional neural networks (CNNs), offer promising tools for automating disease detection based on visual symptoms. Methods: This study employed a sequential CNN algorithm for skin lesion detection using digital images. The dog lesion dataset which consisted of 551 images categorized into five lesions: bumps, hair loss, hot spots, rashes and sores classes, was utilized for training and testing of the CNN model. The outcomes of the proposed model after training and testing were obtained in terms of performance metrics such as recall, accuracy, precision and F1-score. The adaptability and effectiveness of the model in classifying dog skin lesions were accessed by the overall accuracy of positive prediction. Result: The developed model achieved a remarkable overall accuracy rate of 84.38% in classifying lesion classes. Its precision and recall tests validate its dependability, underscoring its potential to improve ailments detection and treatment in dogs, eventually increasing their well-being.