[Paper Review] A Review: Expert System for Diagnosis of Myocardial Infarction

S. J. Gath, R. V. Kulkarni|arXiv (Cornell University)|Jan 1, 2014

Cerebral Palsy and Movement Disorders2 references4 citations

TL;DR

This paper reviews expert systems for diagnosing myocardial infarction, focusing on how they manage uncertainty in imprecise, incomplete, or unreliable medical data. It synthesizes past research on AI-driven cardiac diagnosis systems, emphasizing knowledge representation and reasoning techniques to achieve human-expert-level performance in narrow clinical domains.

ABSTRACT

A computer Program Capable of performing at a human-expert level in a narrow problem domain area is called an expert system. Management of uncertainty is an intrinsically important issue in the design of expert systems because much of the information in the knowledge base of a typical expert system is imprecise, incomplete or not totally reliable. In this paper, the author present s the review of past work that has been carried out by various researchers based on development of expert systems for the diagnosis of cardiac disease

Motivation & Objective

• To analyze the evolution and design principles of expert systems in diagnosing myocardial infarction.
• To examine how uncertainty management is addressed in cardiac diagnostic expert systems.
• To evaluate the integration of clinical knowledge and AI techniques for improved diagnostic accuracy.
• To identify gaps and limitations in existing systems based on prior research.
• To provide a foundation for future development of robust, reliable expert systems in cardiology.

Proposed method

• Systematic review of published research on expert systems for myocardial infarction diagnosis.
• Analysis of knowledge representation techniques used in clinical expert systems.
• Evaluation of uncertainty handling mechanisms such as probabilistic reasoning and fuzzy logic.
• Comparison of rule-based systems, decision trees, and hybrid models in cardiac diagnosis.
• Synthesis of findings from multiple studies to identify common design patterns and challenges.
• Focus on systems that aim for human-expert-level performance in narrow diagnostic domains.

Experimental results

Research questions

• RQ1 How do expert systems model and manage uncertainty in cardiac diagnostic data?
• RQ2 What are the key knowledge representation and reasoning techniques used in myocardial infarction diagnosis systems?
• RQ3 How do these systems compare in performance to human cardiologists in specific diagnostic tasks?
• RQ4 What are the limitations of current expert systems in handling incomplete or imprecise clinical data?
• RQ5 What design patterns emerge across different expert systems for cardiac disease diagnosis?

Key findings

• Expert systems for myocardial infarction diagnosis rely heavily on rule-based reasoning and uncertainty management techniques.
• Many systems incorporate probabilistic or fuzzy logic to handle imprecise clinical data such as ECG readings and symptom reports.
• The integration of clinical guidelines and medical expertise into knowledge bases is critical for system reliability.
• Despite progress, challenges remain in handling data incompleteness and variability in real-world clinical settings.
• The review identifies a need for more standardized evaluation and benchmarking of such systems.
• Overlap with prior work (e.g., arXiv:1006.4544) suggests ongoing development in this research area with evolving methodologies.