Detection of Seborrheic & Atopic Dermatitis, Psoriasis, Onychomycosis Diseases with Artificial Intelligence

Kevser Şahinbaş; Berna Şevval Uysal; Sude Tozmaz; Zeliha Sezgin

Makale detayları

Detection of Seborrheic & Atopic Dermatitis, Psoriasis, Onychomycosis Diseases with Artificial Intelligence

Yazarlar:

Kevser Şahinbaş | Berna Şevval Uysal | Sude Tozmaz | Zeliha Sezgin

Yayıncı:

Giresun University

Ana Başlıklar:

Bilgisayar Yazılımı

Dergi:

Karadeniz Fen Bilimleri Dergisi Vol.15 Issue 3, pp.1004-1026

Belge Tipi:

Makale

Yayın Tarihi:

9/2025

Belge Dili:

Türkçe

Document ID (Org No.):

2001358292

pISSN:

2564-7377

Anahtar kelimeler:

Özet:

 This study aims to accurately and rapidly detect seborrheic dermatitis, atopic dermatitis, psoriasis and onychomycosis diseases using artificial intelligence technology. The use of artificial intelligence technologies in this field offers significant advantages by accelerating diagnosis processes and increasing the accuracy rate. Deep learning techniques, an important branch of artificial intelligence, were used in this study. It is aimed to diagnose skin diseases with high accuracy from images using the YOLO algorithm.The dataset obtained from Kaggle was used to develop the proposed model and the model was trained. The dataset contains high-resolution skin images of seborrheic dermatitis, atopic dermatitis, psoriasis and onychomycosis diseases. In the methodology section of the study, the stages of pre-processing the data, developing the model and testing the model are discussed in detail. Before the data was submitted to machine learning, single-class classification operations were performed via RoboFlow and the images were brought to the standard size of 640x640. For the model, the yolov8s-cls model of the Yolov8 algorithm was used and coded in the Google Colab environment. During the training of the model, the performance of the model was evaluated using training, validation and test data sets. The model reached an accuracy rate of 88.3 and successfully classified the diseases. The accuracy rates obtained reveal the potential of such models to be used in clinical applications. This study aims to contribute to the widespread use of artificial intelligence-based systems in the healthcare sector and to the improvement of diagnostic processes.

 This study aims to accurately and rapidly detect seborrheic dermatitis, atopic dermatitis, psoriasis and onychomycosis diseases using artificial intelligence technology. The use of artificial intelligence technologies in this field offers significant advantages by accelerating diagnosis processes and increasing the accuracy rate. Deep learning techniques, an important branch of artificial intelligence, were used in this study. It is aimed to diagnose skin diseases with high accuracy from images using the YOLO algorithm.The dataset obtained from Kaggle was used to develop the proposed model and the model was trained. The dataset contains high-resolution skin images of seborrheic dermatitis, atopic dermatitis, psoriasis and onychomycosis diseases. In the methodology section of the study, the stages of pre-processing the data, developing the model and testing the model are discussed in detail. Before the data was submitted to machine learning, single-class classification operations were performed via RoboFlow and the images were brought to the standard size of 640x640. For the model, the yolov8s-cls model of the Yolov8 algorithm was used and coded in the Google Colab environment. During the training of the model, the performance of the model was evaluated using training, validation and test data sets. The model reached an accuracy rate of 88.3 and successfully classified the diseases. The accuracy rates obtained reveal the potential of such models to be used in clinical applications. This study aims to contribute to the widespread use of artificial intelligence-based systems in the healthcare sector and to the improvement of diagnostic processes.

 This study aims to accurately and rapidly detect seborrheic dermatitis, atopic dermatitis, psoriasis and onychomycosis diseases using artificial intelligence technology. The use of artificial intelligence technologies in this field offers significant advantages by accelerating diagnosis processes and increasing the accuracy rate. Deep learning techniques, an important branch of artificial intelligence, were used in this study. It is aimed to diagnose skin diseases with high accuracy from images using the YOLO algorithm.The dataset obtained from Kaggle was used to develop the proposed model and the model was trained. The dataset contains high-resolution skin images of seborrheic dermatitis, atopic dermatitis, psoriasis and onychomycosis diseases. In the methodology section of the study, the stages of pre-processing the data, developing the model and testing the model are discussed in detail. Before the data was submitted to machine learning, single-class classification operations were performed via RoboFlow and the images were brought to the standard size of 640x640. For the model, the yolov8s-cls model of the Yolov8 algorithm was used and coded in the Google Colab environment. During the training of the model, the performance of the model was evaluated using training, validation and test data sets. The model reached an accuracy rate of 88.3 and successfully classified the diseases. The accuracy rates obtained reveal the potential of such models to be used in clinical applications. This study aims to contribute to the widespread use of artificial intelligence-based systems in the healthcare sector and to the improvement of diagnostic processes.

 Bu çalışma, seboreik dermatit, atopik dermatit, psoriasis ve onikomikoz hastalıklarının yapay zekâ teknolojisi kullanılarak doğru ve hızlı bir şekilde tespit edilmesini hedeflemektedir. Yapay zekâ teknolojilerinin bu alanda kullanımı, teşhis süreçlerini hızlandırarak ve doğruluk oranını artırarak önemli avantajlar sunmaktadır. Bu çalışma kapsamında, yapay zekânın önemli bir dalı olan derin öğrenme teknikleri kullanılmıştır. YOLO algoritması kullanılarak, cilt hastalıklarının görüntüler üzerinden yüksek doğrulukla teşhis edilmesi amaçlanmıştır. Önerilen modelin geliştirilmesi için Kaggledan elde edilen veri seti kullanılmış ve model eğitilmiştir. Veri seti, seboreik dermatit, atopik dermatit, psoriasis ve onikomikoz hastalıklarına ait yüksek çözünürlüklü cilt görüntülerini içermektedir. Çalışmanın metodoloji kısmında, verilerin ön işleme tabi tutulması, modelin geliştirilmesi ve test edilmesi aşamaları ayrıntılı bir şekilde ele alınmıştır. Veriler makine öğrenimine sunulmadan önce RoboFlow üzerinden tek sınıflı sınıflandırma işlemleri yapılmış ve görüntüler 640x640 standart boyutlarına getirilmiştir. Model için, Yolov8 algoritmasının yolov8s-cls modeli kullanılmış ve Google Colab ortamında kodlanmıştır. Modelin eğitimi sırasında, eğitim, doğrulama ve test veri setleri kullanılarak modelin performansı değerlendirilmiştir. Model, 88.3 doğruluk oranına ulaşmış ve hastalıkları başarılı bir şekilde sınıflandırmıştır. Elde edilen doğruluk oranları, bu tür modellerin klinik uygulamalarda kullanılma potansiyelini ortaya koymaktadır. Bu çalışma, sağlık sektöründe yapay zekâ tabanlı sistemlerin kullanımının yaygınlaşmasına ve teşhis süreçlerinin iyileştirilmesine katkıda bulunmayı amaçlamaktadır.

Makale detayları

Detection of Seborrheic & Atopic Dermatitis, Psoriasis, Onychomycosis Diseases with Artificial Intelligence

Yazarlar:

Kevser Şahinbaş | Berna Şevval Uysal | Sude Tozmaz | Zeliha Sezgin

Yayıncı:

Giresun University

Ana Başlıklar:

Bilgisayar Yazılımı

Dergi:

Karadeniz Fen Bilimleri Dergisi Vol.15 Issue 3, pp.1004-1026

Belge Tipi:

Makale

Yayın Tarihi:

9/2025

Belge Dili:

Türkçe

Document ID (Org No.):

2001358292

pISSN:

2564-7377

Skin Diseases

Artificial Intelligence

RoboFlow

YOLOv8

Classification

Convolutional Neural NetworksCilt Hastalıkları

Yapay Zekâ

RoboFlow

YOLOv8

Sınıflandırma

Evrişimsel Sinir Ağları

Sayı

Makale detayları

Yazarlar: Kevser Şahinbaş | Berna Şevval Uysal | Sude Tozmaz | Zeliha Sezgin

Yayıncı: Giresun University

Ana Başlıklar: Bilgisayar Yazılımı

Dergi: Karadeniz Fen Bilimleri Dergisi Vol.15 Issue 3, pp.1004-1026

Belge Tipi: Makale

Yayın Tarihi: 9/2025

Belge Dili: Türkçe

Document ID (Org No.): 2001358292

pISSN: 2564-7377

Sayı

Yazarlar:

Kevser Şahinbaş | Berna Şevval Uysal | Sude Tozmaz | Zeliha Sezgin

Yayıncı:

Giresun University

Ana Başlıklar:

Bilgisayar Yazılımı

Dergi:

Karadeniz Fen Bilimleri Dergisi Vol.15 Issue 3, pp.1004-1026

Belge Tipi:

Makale

Yayın Tarihi:

9/2025

Belge Dili:

Türkçe

Document ID (Org No.):

2001358292

pISSN:

2564-7377