A scalable multi-agent architecture in environments with limited connectivity: Case study on individualised care for healthy pregnancy

Technology advancement has motivated researches and studies in different domains which have enormous impact on the quality of life including healthcare domain. Among the challenges that hinder effective realization of these technologies in healthcare domain are heterogeneous nature of patients, limited internet connectivity - especially in developing countries -, and high costs of healthcare services resulting from hospitalizations and treatments of critical health conditions. In this paper we have addressed the solution to these challenges by using a case study of pregnant women. We firstly present the current status in the management of pregnancy complications, which motivates the need for improvement. We then describe analysis and design models by following agent-oriented modelling, which considers man-made agents (software agents) and human agents. Finally we present and discuss multi-agent architecture which addresses a solution to the outlined above healthcare challenges by considering (1) scalability of the provided healthcare service in environments with limited connectivity (2) providing patient care in accordance with the characteristics of individual patients such as allergies, medical history and hobbies (3) prediction of critical health conditions by using multi-parametric machine learning algorithms, which aim to provide early diagnosis of critical health condition, and (4) providing home care that enables patients to collect their physiological data and submit them to the hospital information system for continuous monitoring and analysis.