Telemedicine System Supporting Early Diagnosis and Efficient Therapy of Lyme Disease

Abstract

This work presents a telemedicine system designed to support the diagnosis of Lyme disease, a common and dangerous illness spread by ticks. The system was designed as a smartphone application and built in close cooperation with doctors specializing in diagnosing and treating Lyme disease. After logging in, a potential patient answers yes/no to a series of simple questions in a properly composed survey. Then, he is asked to take a photo of the skin lesion (erythema migrans) with a smartphone. The complete data set is sent to the contractual system administrator, and artificial intelligence performs a preliminary data analysis. As a result, the patient is sent information that the probability of the disease is low or high. The system advises seeing a specialist in high-risk cases for a complete diagnosis and treatment. Ignoring early symptoms can lead to severe complications in the later stages of the disease. The paper presents preliminary results of diagnoses made by neural networks. Despite being conducted on a small dataset, the research showed promising results with 93% accuracy. The conclusion highlights the system's practical applications and potential for similar uses.