Performance of the CNN Method for Identifying Health Conditions Based on Nail Images

Abstract

Generally, someone who feels sick in his body wants to know his actual health condition by visiting a doctor at a hospital or other health service center for a medical examination. The doctor will ask about the symptoms experienced by the patient and sometimes examine several parts of the body to get important information before he diagnoses the patient's health condition. Theoretically, based on research developments in the medical field, changes in conditions (related to color, texture, and shape) on fingernails or toenails indicate changes in the health condition of a person's body. When someone has a health problem, the body's nerves will send signals to the fingernail or toenail area, and then the physical condition of the nail changes color, texture, and shape. These changes occur slowly, according to the condition of a person's body. Each type of health disorder or disease in the body will produce unique nail changes. Visually, the physical changes of the nails are often not very visible, but if you look closely, these changes do occur. Our research proposes an intelligent system (an artificial intelligence-based software application) to automatically diagnose body health conditions using photos of fingernails. The analysis process is carried out on the nail image to find out whether someone has health problems or not. The system for detecting body health conditions automatically using photos of nails produced in our research has a relatively good performance, namely an accuracy of 86.45%, a precision of 0.78, a recall of 1.0, and an f1 score of 0.88.