Medical diagnostic system using Fuzzy Coloured Petri Nets under uncertainty

Area of Science:

• Computer Science
• Artificial Intelligence
• Medical Informatics

Background:

• Traditional Petri Nets require extensive subnets for similar processes, leading to large, unmanageable models.
• Existing machine learning methods for knowledge-based systems have a significant gap between knowledge-intensive and knowledge-free approaches.
• Fuzzy Petri Nets (FPN) offer fuzzy reasoning but can still result in large models for complex knowledge.

Purpose of the Study:

• To propose a hybrid learning and reasoning method using Fuzzy Coloured Petri Nets (FCPN) to handle uncertainty in medical diagnostics.
• To address the issue of large, repetitive subnets in traditional Petri Net models.
• To integrate knowledge-based FCPN with Genetic Algorithms (GA) for a more effective diagnostic system.

Main Methods:

• Development of a learning and reasoning method based on Fuzzy Coloured Petri Nets (FCPN) for uncertain environments.
• Implementation of a hybrid learning approach combining knowledge-based FCPN with Genetic Algorithms (GA).
• Application of the proposed system to the specific domain of intervertebral disease diagnosis.

Main Results:

• The proposed FCPN-based system effectively performs fuzzy reasoning and learning under uncertainty.
• The hybrid FCPN and GA method successfully overcomes the limitations of large, repetitive subnet structures.
• The system demonstrated validity and effectiveness in diagnosing intervertebral diseases.

Conclusions:

• The hybrid FCPN and GA approach provides an efficient and scalable solution for medical diagnostic systems.
• This method enhances reasoning capabilities in uncertain conditions, improving diagnostic accuracy.
• The successful application to intervertebral diseases highlights the potential of this approach for complex medical diagnoses.