This paper presents an assessment of liver function using novel neuro-fuzzy model of blood albumin level (BA). The developed model that is used to predict the BA consists of four inputs: Asparate Aminotransferace (AST), Alkaline Phosphate (ALP), Total Bilirubin (T. Bil.) and Total Protein (T. Prot.), which are measured in any routine liver function test. The proposed BA model was trained using 211 measured data and a root-mean square error (RMSE) of 0.29 for 100 epochs was achieved. The performance of the developed BA model was validated using 57 testing data sets and RMSE of 0.34 for 100 epochs was achieved. The correlation coefficient (CC) between the predicted and measured values of blood albumin is statistically significant (CC=0.83), which ensures the efficiency and accuracy of developed fuzzy model for predicting BA. The main clinical benefit of this model is that it improves the assessment capabilities of liver diseases and can be used as an integral part of any medical expert system denoted for assessment and diagnosis of liver disorders.
Top1. Introduction
Liver is a complex biological organ that has great biochemical functions. These include removing of toxins and drugs from blood stream, regulation of glucose level, fats, hormones and amino acids, and also storing vitamins and iron (Sullivan, 2001; Rodoman, Shaeva, Dobretsov, & Korotaev, 2006; Kratz et al., 2004). Clinically, the assessment of these functions is carried out by making use of different tests (Cales & Oberti, 2002; Blum, 1998) performed on patient’s blood. One of these tests is the albumin test because the albumin is the main protein in blood synthesized by the liver. It provides indication of the ability of the liver to synthesize proteins and it is used to assess the degree of liver damage. Therefore, with normal liver function, the concentration of blood albumin is also normal. Thus, an abnormally elevated level of BA is an efficient measure of impaired liver function and damage of liver tissues. Furthermore, the blood albumin functions as a transport protein and also maintains the osmotic pressure between blood and biological tissues. Thus, accurate measurement or prediction (indirect measurement) of BA has great clinical importance for evaluation of liver function and tissues damage.
The concentration of blood albumin is measured by making use of immunological (IM), dye-binding (DB), and electrophoretic (EL) and fluorescence immunoassay (FI) methods. Clinically, the dye-binding methods, bromcresol green (BCG) and bromcresol purple (BCP) are routine methods for BA measurements because they have good sensitivity (Evans & Parsons, 1988; Gustafsson, 1976). However, the use of either BCG or BCP can yield falsed elevated values of BA because they tend to overestimate the blood albumin especially when the BA is low (Hall, 1992; Bush & Reeds, 1987). Further, it consumes large time to conduct the BA measurement and requires well qualified operator to handle the measurement. The IM methods have high specificity but unfortunately they have many limitations (Marre, Clandel, Ciret, Luis, Snarez, & Passa, 1987). First, they need a centrifugation of blood sample in order to carry out the albumin measurement. Second, they require long analysis time to measure the BA. Finally, the IM methods are more expensive than the dye-binding methods. The EL methods have good sensitivity and require short analysis time but again they need more improvements in terms of preparation of blood sample and measuring procedure (Miki, Kaneta, & Imasaka, 2001). Although, the FI methods do not require a centrifugation of blood sample, they still inconvenient because they are time consuming (Brooks, Devine, Harris, Harris, Miller, & Olal, 1999).