MATLAB-Based Implementation of a Multiclass Dermatological Diagnostic Model Using Convolutional Neural Networks and Preprocessing-Optimized Image Features.
- Version
- Download 13
- File Size 389.20 KB
- File Count 1
- Create Date 25 July 2025
- Last Updated 25 July 2025
MATLAB-Based Implementation of a Multiclass Dermatological Diagnostic Model Using Convolutional Neural Networks and Preprocessing-Optimized Image Features.
Sahil Rajadhyaksha1, Pranali Baviskar2, Preksha Koli3
1Department of Electronics and Telecommunication, Vidyalankar Institute of Technology.
2Department of Electronics and Telecommunication, Vidyalankar Institute of Technology.
3Department of Biomedical Engineering, Vidyalankar Institute of Technology.
Abstract - Skin diseases such as melanoma, vitiligo, psoriasis, eczema, and acne are among the most prevalent globally, often requiring timely diagnosis to avoid complications. Traditional diagnostic methods are highly dependent on clinical expertise and visual inspection, which can be subjective and time-consuming. In this paper, we present a multiclass skin disease classifier built entirely in MATLAB, integrating image preprocessing with a pretrained Convolutional Neural Network through a custom-designed graphical user interface. The system can classify five common skin conditions like eczema, melanoma, ringworm, acne, and vitiligo based on dermatological images. The integrated GUI simplifies the diagnostic process, making the tool suitable for non- technical users such as clinicians and medical students. This system is particularly promising for deployment in low-resource or remote settings where access to dermatological expertise is limited.
Key Words: Skin Disease Detection, Convolutional Neural Network, MATLAB, Image Processing, Medical Diagnosis, GUI, Multiclass Classification