Skip to Main content Skip to Navigation

Approche Intelligence Artificielle Distribuée pour une planification réactive et une aide à la conduite du processus de blocs opératoires hospitaliers.

Abstract : The operating theater is one of the most important sector of a hospital. Optimizing its process is a priority for hospital managers. The abundant literature on the subject is unanimous on the fact that planning and sequencing interventions, satisfying a great multitude of requirements and constraints makes the construction of the operational program a very complex task. In addition, the construction of the program is far from enough, it is imperative to maintain automatic in "Near Real Time" the schedule according to the events appearing during the downstream phase of realization.The objective of this thesis is to provide a methodology for the management of the operating room process, integrating on the one hand a decision support to optimize the predictive planning and on the other hand to allow a dynamic replanning to guarantee a reactivity of the process of this medical sector.Our study began with a state of the art on the issues of operating theater management and proposed solutions, with an initial goal: to target the improvement approach to engage. We have noticed that the complexity and the heaviness of the existing models often based on the operational research and especially the combinatorial explosion of the constraints makes it impossible to find the optimum for realistic sizes of instances. This led us to decide that our study would break with the traditional strategy of continuous improvement. We favored an IAD approach aimed at finding solutions according to a programmed logic rather than the calculated resolutions of traditional approaches. Subsequently, we modeled the actors in the surgical intervention planning process using a multi-agent system.The proposed approach provides the most appropriate predictive planning of the operating theater activity. Each surgical procedure is planned individually considering the rules, all the constraints but also the surgeons' preferences. The same model allows in the downstream phase with a decision aid, the assignment of a surgical intervention facing an emergency situation. For a better performance coupled with an increased reactivity, we have also approached, with this model, the problem of dynamic adjustment of the schedules to balance the load in hours of surgery between the operating rooms.In the proposed methodology, decision support is based on the use of knowledge and the rules of know-how and their capitalization according to the principles of an expert system. In addition, we have proposed a set of automatic calculated performance and responsiveness indicators that are likely to be implemented in a future validation tool and that can be used as feedback to optimize the decision support process.Collaboration with a CHU in Lebanon provided data on scenarios of weekly surgeries. The results of some simulations highlight the convenience of the IAD approach in solving the problem.
Complete list of metadatas

Cited literature [281 references]  Display  Hide  Download
Contributor : Abes Star :  Contact
Submitted on : Friday, June 5, 2020 - 6:03:08 PM
Last modification on : Tuesday, October 27, 2020 - 2:34:45 PM


Version validated by the jury (STAR)


  • HAL Id : tel-02799649, version 1



Bilal Bou Saleh. Approche Intelligence Artificielle Distribuée pour une planification réactive et une aide à la conduite du processus de blocs opératoires hospitaliers.. Intelligence artificielle [cs.AI]. Université Bourgogne Franche-Comté; Université libanaise, 2019. Français. ⟨NNT : 2019UBFCA028⟩. ⟨tel-02799649⟩



Record views


Files downloads