Enriching Dermatological Disease Detection Through Deep Learning

  • Anandkumar Rao, Dr. Geetanjali V. Kale

Abstract

One of the most effective first line of defense against foreign objects and germs for getting into the body is the skin. The skin is the biggest and one of the most defensive organs in the body. Skin illnesses are quite prevalent and come in a variety of forms as a result of being perpetually targeted. As a result, detecting skin disorders is a highly challenging undertaking. Skin disease diagnosis is a complex procedure carried out by doctors and experts using a variety of tests and exams. This is a time-consuming technique that can occasionally result in false positives and can be troublesome because it consumes a significant amount of the practitioner’s time. There is a growing curiosity in the remote and automated diagnosis methodology, which can be an effective implementation in this field. As a result, an effective strategy for a skin identification technique based on machine learning and image processing implementations has been suggested. The presented approach consists of image normalization, the YCbCr model, Convolutional Neural Networks, and a Decision Tree. The technique has been effective in improving skin disease identification, and the experimental findings have achieved Precision, recall and F Measure Values of 74.76%, 74.76%, and 74.54% respectively.

Published
2021-10-12
How to Cite
Anandkumar Rao, Dr. Geetanjali V. Kale. (2021). Enriching Dermatological Disease Detection Through Deep Learning. Design Engineering, 3280 - 3291. https://doi.org/10.17762/de.vi.5269
Section
Articles