Publication Date: 2023/11/07
Abstract: Dermatological diseases are highly prevalent and affect individuals of all ages and genders. Accurate prediction of these diseases is crucial for timely diagnosis and effective treatment. Skin lesions, characterized by variations in color, shape, and texture, serve as important indicators of dermatological conditions. In this research, we have conducted a comparative analysis of different models to detect and recognize skin diseases. The objective of our study is to develop a model that can accurately predict various dermatological diseases. The importance of our research lies in addressing the widespread nature of dermatological diseases and the need for reliable and efficient prediction methods. By employing machine learning techniques, we aim to provide a tool that can assist dermatologists in their diagnosis and decision-making processes. To determine the most effective approach, we evaluated the performance of various models. Among them, densenet121 demonstrated the highest accuracy and reliability. We got an accuracy of 90.6% using this model. Therefore, we selected densenet121 as the basis for our proposed method. By implementing the densenet121 model, we achieved significant improvements in the prediction of dermatological diseases. Our findings indicate that this model can accurately identify and classify different skin lesions, enabling early detection and timely intervention. In conclusion, our research highlights the significance of accurate prediction models in the field of dermatology. The utilization of densenet121 as a basis for our proposed method shows promising results, emphasizing its potential as an efficient tool for dermatological disease prediction. The development and integration of such models into clinical practice can significantly contribute to improved patient outcomes and enhance the overall management of dermatological conditions.
Keywords: Skin Lesion, Neural Network,, Convolutional Neural Network, DenseNet-121
DOI: https://doi.org/10.5281/zenodo.10078152
PDF: https://ijirst.demo4.arinfotech.co/assets/upload/files/IJISRT23OCT1717.pdf
REFERENCES