This paper presents two novel approaches to determine optimum growing multi-experts network (GMN) structure. The first method called direct method deals with expertise domain and levels in connection with local experts. The growing neural gas (GNG) algorithm is used to cluster the local experts. The concept of error distribution is used to apportion error among the local experts. After reaching the specified size of the network, redundant experts removal algorithm is invoked to prune the size of the network based on the ranking of the experts. However, GMN is not ergonomic due to too many network control parameters. Therefore, a self-regulating GMN (SGMN) algorithm is proposed. SGMN adopts self-adaptive learning rates for gradient-descent learning rules. In addition, SGMN adopts a more rigorous clustering method called fully self-organized simplified adaptive resonance theory in a modified form. Experimental results show SGMN obtains comparative or even better performance than GMN in four benchmark examples, with reduced sensitivity to learning parameters setting. Moreover, both GMN and SGMN outperform the other neural networks and statistical models. The efficacy of SGMN is further justified in three industrial applications and a control problem. It provides consistent results besides holding out a profound potential and promise for building a novel type of nonlinear model consisting of several local linear models.
In this paper, we highlight the involvement of Knowledge Management in a healthcare enterprise. We argue that the 'knowledge quotient' of a healthcare enterprise can be enhanced by procuring diverse facets of knowledge from the seemingly placid healthcare data repositories, and subsequently operationalising the procured knowledge to derive a suite of Strategic Healthcare Decision-Support Services that can impact strategic decision-making, planning and management of the healthcare enterprise. In this paper, we firstly present a reference Knowledge Management environment-a Healthcare Enterprise Memory-with the functionality to acquire, share and operationalise the various modalities of healthcare knowledge. Next, we present the functional and architectural specification of a Strategic Healthcare Decision-Support Services Info-structure, which effectuates a synergy between knowledge procurement (vis-à-vis Data Mining) and knowledge operationalisation (vis-à-vis Knowledge Management) techniques to generate a suite of strategic knowledge-driven decision-support services. In conclusion, we argue that the proposed Healthcare Enterprise Memory is an attempt to rethink the possible sources of leverage to improve healthcare delivery, hereby providing a valuable strategic planning and management resource to healthcare policy makers.
Existing Problem-Based Learning (PBL) problems, though suitable in their own right for teaching purposes, are limited in their potential to evolve by themselves and to create new knowledge. Presently, they are based on textbook examples of past cases and/or cases that have been transcribed by a clinician. In this paper, we present (a) a tacit healthcare knowledge representation formalism called Healthcare Scenarios, (b) the relevance of healthcare scenarios in PBL in healthcare and medicine, (c) a novel PBL-Scenario-based tacit knowledge explication strategy and (d) an online PBL Problem Composer and Presenter (PBL-Online) to facilitate the acquisition and utilisation of expert-quality tacit healthcare knowledge to enrich online PBL. We employ a confluence of healthcare knowledge management tools and Internet technologies to bring tacit healthcare knowledge-enriched PBL to a global and yet more accessible level.
The incidences of breast cancer have been rising at an alarming rate. Mass breast screening programmes involving mammography and ultrasound in certain parts of the world have also proven their benefits in early detection. However, radiologists may be confronted with increased workload. An attempt has been made in this paper to rectify part of the problems faced in this area. Expert systems based on the interpretation of mammographic and ultrasound images for classifying patient cases could be utilized by doctors (expert and non-expert) in screening. These softwares consist of MAMMEX (for mammogram) and SOUNDEX (for breast ultrasound) could be used to deduce cases according to Breast Imaging Recording and Data System (BI-RADS), based on patients’ history, physical and clinical assessment, mammograms and breast ultrasound images. A total of 179 retrospective cases from the Radiology Department, hospital of the University of Science Malaysia, Kubang Kerian, Kelantan were used in this study. A receiver operating characteristic (ROC) curve analysis was implemented, based on the usage of a two-class forced choice of classifying suspicious and malignant findings as positive with normal, benign and probably benign classified as negative. Results yielded an area under the curve (AUC) of 0.997 with the least standard error value of 0.003 for MAMMEX while an AUC of 0.996 with the least standard error of 0.004 was accomplished for SOUNDEX. A system which very closely simulated radiologists was also successfully developed in this study. The ROC curve analysis indicated that the expert systems developed were of high performance and reliability.