Rough set based lattice structure for knowledge representation in medical expert systems: Low back pain management case study. (1st May 2020)

• Record Type: Journal Article
• Authors: Santra, Debarpita; Basu, Swapan Kumar; Mandal, Jyotsna Kumar; Goswami, Subrata
• Is Part Of: Expert systems with applications. Volume 145(2020)
• Publication Date: 2020-05-01
• Subjects: Medical expert system -- Lattice structure -- Rough set theory -- Knowledge-base -- Low back pain; Expert systems (Computer science) -- Periodicals; Systèmes experts (Informatique) -- Périodiques; Electronic journals; 006.33
• DOI: 10.1016/j.eswa.2019.113084
• Languages: English
• ISSNs: 0957-4174

Abstract

Highlights: Exhaustive but non-redundant medical knowledge acquisition by lattice-based approach. Easy modification of knowledge base maintaining completeness and consistency. Measured reliability of decision rule and re-defined certainty and coverage factor. Achieved computational efficiency using rough-lattice based knowledge representation. Generated optimal rule sets to be used directly in clinical diagnosis. Abstract: The aim of medical knowledge representation is to capture detailed domain knowledge in a clinically efficient manner, and to offer reliable resolution with the acquired knowledge. The knowledge base should allow incremental growth with inclusion of updated knowledge over the time. Accommodating the knowledge acquired from a variety of knowledge sources by different knowledge engineers may lead to a redundant and inconsistent design of the knowledge base, increasing the storage size and the time for knowledge retrieval. In this paper, we have proposed a rough set based lattice framework for representation of knowledge in medical expert systems which overcomes the problem of redundancy and inconsistency in knowledge and offers computational efficiency with respect to both time and space. The proposed knowledge representation offers a flexible scheme for expressing diverse possibilities of inter-relatedness among the symptoms and diseases in a systematic manner within the lattice structures. Through our design, we generate an optimal set of decision rules for use during inference. Reliability of each rule is measured using a new metric called credibility factor, and the certainty and coverage factors of a decision rule have been re-defined. During inference, the medical expert system considers the highly reliable and certain rules first, and the possible and uncertain rules at the later stages, if recommended by physicians. The proposed scheme ensures completeness, consistency, integrity, non-redundancy, and ease of access. These qualities are automatically preserved in the designed knowledge base while being updated with the new knowledge of medical advancements. As a result, the overall maintenance cost of a medical expert system is significantly reduced. The proposed knowledge representation technique has been illustrated using an example from the domain of low back pain. Though the proposed technique has been demonstrated for low back pain, it offers a generalized scheme for knowledge representation and may be adopted by medical expert systems (e.g. pathology, psychology, and many other specialties) having complexity similar to this one; no application-specific knowledge representation technique needs to be designed. The medical expert system developed using this scheme may act as a standardized example for development of different unexplored or semi-explored health care automated systems. This scheme may also be applied for pattern recognition, rule mining, and conflict analysis in complex medical data domains. Our design will have a wide ranging impact towards finding low-cost, reliable and available healthcare solutions for different diseases, especially in primary care units where expert physicians are scarce.