ASSESSMENT OF WEIGHT CHANGES AND SUGGESTION OF NUTRITIONAL FORMULAS FOR PREMATURELY BORN BABIES WITH CLOSEST REASONABLE CENTROIDS

Abstract

Recently, many models of applying artificial intelligence (AI) techniques into the analysis of clinical data have been proposed. Unfortunately, most models provide little help when specific "cause–effect" relation of data is not available, or even known. In this paper, an innovative method, called closest reasonable centroids (CRC), is directed to address this issue. Our present application domain was a clinical data set of the weight changes of 274 prematurely born babies who had nutritional deficiency problem and were given TPN treatments to improve their nutritional needs. Experimental result shows that the CRC's differentiability is comparable to that of the back-propagation neural networks (BPN) and better than that of statistical method. Also, from the health conditions of babies and their nutritional treatments, the proposed method can roughly predict their weight changes and provide some suggested feasible formula. All of the above results have been double confirmed by the clinicians, implicating that CRC could be used as assistant tool.