Article Preview
TopBackground
Liver is the largest glandular organ of the human body, which weighs around three pounds (Li et al., 2012). The liver performs different types of metabolic functions, like filtering blood, producing bile, assisting in fat digestion, making proteins for blood clotting, metabolising drugs, storing glucose and, most importantly, detoxifying harmful chemicals Singh et al. (2017). Malfunctioning of liver may cause liver disease and have serious health effects. The causes of liver disease are varied and can include consumption of contaminated food, inherited disorders, accumulation of excessive fat, hepatocytes damage due to infection by bacteria, viruses or fungi, and excessive consumption of alcohol or drugs Lin et al. (2010). Malnutrition, obesity leads to advanced stage of liver disease and non-alcoholic fatty liver disease further leading to non-alcoholic steatohepatitis and cirrhosis for some cases as discussed by McClain et al. (2020). They have also discussed various causes and the methods for assessing malnutrition. A patient’s survival rate may increase by several times if the diagnosis of the liver disease is done at an early stage, but diagnosis requires various examination tests by expert physicians. However, these do not always assure the correct diagnosis Takkar et. al (2017), but liver function tests do significantly help in examining liver disorders. The key parameters in these tests include albumin, alkaline phosphatase, total proteins, alanine aminotransferase, aspartate aminotransferase, direct bilirubin, total bilirubin, gamma-glutamyl transferase, prothrombin time, triglycerides and platelet counts. Liver diseases are categorised into more than 100 types, and the disease can be acute or chronic. Some liver disease has successful treatments, while others do not.
Liver disease and prediction now increasingly depend on intelligent systems, which now play a significant role in the medical industry. Data mining algorithms, neural networks and statistical techniques are widely applied to liver examination data to evaluate illnesses. Predictive modelling is a broadly used intelligent technique for automated detection of multiple diseases. Machine learning calculations provide specialists with essential measurements, continuous information and progressive examination data about a patient’s illness, lab test results, preliminary clinical information, and family history. As identified by Jesty (2019) machine learning tasks in the field of medicine can be classified into categories namely (i) genomics which is the study of DNA, (ii) audio analysis that is interpretable pattern of digital audio recording, (iii) computer vision which extracts information from digital images and videos by running algorithms, (iv) natural language processing which process text for meaningful information and (v) health record regression that establishes relations between features. The guarantee for improving the detection and prediction of disease has increased interest in machine learning in the biomedical field and had improved the decision-making process by increasing its objectivity. Not only for liver but also for any disease the diagnosis involves many uncertainties in the information system and to handle them different types of intelligent techniques are used with a proposed model as shown used by P. and Acharjya (2020).