Agent-based systems

Agent-based systems deploy entities acting with a degree of autonomy in an environment, whilst sensing changes in that environment. The advantage of such a system, is that tasks with a degree of unpredictability can be completed in a robust and scalable way. There are a wide variety of agent systems, varying across architecture (symbolic, reactive), knowledge representation, reasoning (deductive, means-end), and coordination mechanisms. Agent-based systems have proved to be a successful approach in Industry 4.0. In this post, I will examine if agent-based systems have achieved progress in the domain of healthcare.

The factors driving development of agent-based systems across domains, are complex and dynamic environments, with ‘big data’ (volume, velocity, variety). In such an environment, simple decision-making models are not robust. With the increase in computational power available, including parallel processing, development of intelligent agents has been possible.

In one ambitious project, ‘Agent Hospital’ (Kirn et al., 2006), multiple multi-agent systems were developed, but not yet deployed. For example, the ‘Med Page’ project uses agents representing patients that negotiate autonomously with each other for limited hospital resources, implementing a market mechanism. In another project, agent negotiate with each other autonomously in booking operative theatre time in a schedule, to efficiently allocate the resource. Quite clearly, the technology readiness level is low. In other works, triage of emergency patients using was simulated using nine agents, with some improvement in waiting time (Rahmat et al., 2013), and a four-agent system (assistant, data mining, manager, monitoring) was used to associated drugs to possible side effects, using a knowledge base and data mining (Ji et al., 2012).

In conclusion, the current state of intelligent agents in healthcare is that systems are either used as simulation to understand the real world or are in early stages of development (pre-deployment) (Iqbal et al., 2016). However, given the high complexity of the environment, agent-based systems have a high potential to improve processes.

References

Iqbal, S., Altaf, W., Aslam, M., Mahmood, W. & Khan, M.U.G. (2016) Application of intelligent agents in health-care: review. Artificial Intelligence Review 46 (1): 83–112. Ji, Y., Shen, F. & Tran, J. (2012) A High Performance Agent-Based System for Reporting Suspected Adverse Drug ReactionsIn: 2012 Ninth International Conference on Information Technology - New Generations2012 Ninth International Conference on Information Technology - New Generations 490–495. Available from: https://ieeexplore.ieee.org/document/6209200 [Accessed 14 February 2024]. Kirn, S., Anhalt, C., Krcmar, H. & Schweiger, A. (2006) Agent. Hospital—health care applications of intelligent agents. Multiagent Engineering: Theory and Applications in Enterprises 199–220. Rahmat, M.H., Annamalai, M., Halim, S.A. & Ahmad, R. (2013) Agent-based modelling and simulation of emergency department re-triage. 2013 IEEE Business Engineering and Industrial Applications Colloquium (BEIAC)2013 IEEE Business Engineering and Industrial Applications Colloquium (BEIAC) 219–224.