Skin Lesion Disease Severity Levels Detection using Convolutional Neural Networks

This study aims to detect the severity levels of Skin Lesion diseases. The study classifies the six different severity levels. Skin, a vital organ, acts as a barrier to external environments. Early detection of skin diseases can improve mortality rates and potentially prevent skin cancer. This early diagnosis is a promising experiment due to its similar appearance. A study was led in the nearby areas of Punjab and Haryana. A complete number of 4,000 pictures has been gathered by utilizing essential and optional assets. The collected data is pre-processed by using data cleaning, data quality evaluation, and data change. Six convolutional layers are utilized in the proposed CNN model. The CNN model recognizes the highlights for infection order. The range of the Fl-Score is from 80.20% of the healthy class to 91.24% The macro average of the F'l-Score is 85.19%. The range of the precision for this confusion matrix is 80.04 to 91.98. A deep learning test is completed by using the Python library brings about a 96% recognition precision. This research study presents a conceptual and thorough analysis of recent research on deep learning-based skin disease diagnosis. Considering the popularity of deep learning, there are still many potential and significant hurdles in the field.