| Autor: |
Alnuaimi, Meena N., Alqahtani, Nourah S., Gollapalli, Mohammed, Rahman, Atta, Alahmadi, Alaa, Bakry, Aghiad, Youldash, Mustafa, Alkhulaifi, Dania, Ahmed, Rashad, Al-Musallam, Hesham |
| Předmět: |
|
| Zdroj: |
CMES-Computer Modeling in Engineering & Sciences; 2024, Vol. 141 Issue 3, p2609-2623, 15p |
| Abstrakt: |
Human beings are often affected by a wide range of skin diseases, which can be attributed to genetic factors and environmental influences, such as exposure to sunshine with ultraviolet (UV) rays. If left untreated, these diseases can have severe consequences and spread, especially among children. Early detection is crucial to prevent their spread and improve a patient's chances of recovery. Dermatology, the branch of medicine dealing with skin diseases, faces challenges in accurately diagnosing these conditions due to the difficulty in identifying and distinguishing between different diseases based on their appearance, type of skin, and others. This study presents a method for detecting skin diseases using Deep Learning (DL), focusing on the most common diseases affecting children in Saudi Arabia due to the high UV value in most of the year, especially in the summer. The method utilizes various Convolutional Neural Network (CNN) architectures to classify skin conditions such as eczema, psoriasis, and ringworm. The proposed method demonstrates high accuracy rates of 99.99% and 97% using famous and effective transfer learning models MobileNet and DenseNet121, respectively. This illustrates the potential of DL in automating the detection of skin diseases and offers a promising approach for early diagnosis and treatment. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
