Automated Computation of Therapies Using Failure Mode and Effects Analysis in the Medical Domain

TL;DR

This work addresses the lack of automated reasoning in FMEA-driven medical decision making by formalizing a framework that converts FMEA models into Markov Decision Processes. It introduces an extended FMEA with variables attached to functions and a qualitative causal graph to compute state transitions under actions, enabling automated planning and action selection. By solving the resulting MDP with standard solvers, the approach derives patient-specific therapies as policies and provides an algorithm to extract the best therapy from the initial state. The authors discuss limitations such as scalability and potential extensions to reinforcement learning and integration with language models for training data and validation, highlighting practical implications for automated, evidence-based medical decision support.

Abstract

Failure mode and effects analysis (FMEA) is a systematic approach to identify and analyse potential failures and their effects in a system or process. The FMEA approach, however, requires domain experts to manually analyse the FMEA model to derive risk-reducing actions that should be applied. In this paper, we provide a formal framework to allow for automatic planning and acting in FMEA models. More specifically, we cast the FMEA model into a Markov decision process which can then be solved by existing solvers. We show that the FMEA approach can not only be used to support medical experts during the modelling process but also to automatically derive optimal therapies for the treatment of patients.

Theorems & Definitions (1)

• definition thmcounterdefinition: fmea Model