Predicting the Status of Thyroid and Cardiovascular Patients According to Their Electronic Records Using Temporal Elements Based on the Combination of Shuffled Frog Leaping Algorithm (SFLA) and Deep Learning

Abstract

Health and treatment are two of the most important application fields of information technology, in which the problem of predicting a disease is highly important. The physician makes such predictions based on the clinical condition of the patient and the level of facilities and advances in medical knowledge for the patient—information technology benefits from multiple methods to help this field. Accordingly, the patient information storage system, drug information, treatment and surgery systems, treatment followup systems, remote treatment systems, etc., aim to facilitate the treatment process. The patient can receive the best services within the shortest time due to these systems and information availability. The doctor can provide services to his patient anywhere in the world. This paper provided a model to predict the condition of patients based on their electronic records using temporal elements based on combining the shuffled frog leaping algorithm (SFLA) and deep learning. Accordingly, the evolutionary shuffled frog leaping algorithm (SFLA) and deep learning were used for preprocessing, feature selection, and classification. Two datasets of cardiovascular and thyroid diseases were utilized in the simulation section to ensure the efficiency of the proposed method. Based on this simulation, the proposed method indicated improvement compared to similar methods in the evaluated datasets. In the cardiovascular diseases dataset, this improvement was recorded as 1.4% and 3.2% compared to the author's previous and updated similar methods, respectively.