Abstract:
Skin diseases are known to be a common disease in Sri Lanka and worldwide, which seriously
affects patient’s life and well-being. However, patients who are suffering from these skin diseases
can be cured if diagnosed and treated in early stages. Due to various factors like delay in seeking
medical assistance, issues in clinical attendance patterns, busy schedules of healthcare institutions
and unavailability of experienced medical resources leads diagnostic issues in both sides of
patients as well as Dermatologists.
Dermatologists often face doubts when diagnosing skin diseases. Such problems may emerge
when comparable or rare skin infections are experienced. In such circumstances, a subsequent
sentiment is highly invited.
Improvements in AI has helped specialists in tackle the issue of diagnosing skin diseases. This
will help specialists and medical care laborers with recognizing skin diseases when required.
However, the available systems can be further improved to a greater extent with the intension of
optimizing the efficiency and accuracy. The aim of this research is to design, develop and evaluate
a tool that can diagnose skin diseases as early as possible. This project presents an intelligent
system to diagnose more than five skin diseases which are common in Sri Lanka, using image
processing and Deep Learning with high accuracy to serve the healthcare sector in Sri Lanka.
Derma Care is a solution to bridge the gap between the public and health care services with the
use of a skin disease diagnosis system, which addresses the common diagnostic issues faced by
both Patients and Dermatologists in Sri Lanka. The methodology focuses on training a
classification model using Convolutional Neural Network. The solution provided by Derma Care
skin disease diagnosis web application was evaluated by the healthcare domain experts, technical
domain experts as well as non-experts who are considered as the users of the application to
identify the practicality and validity of the project.
"
