A New Clinical Spectrum for the Assessment of Nonalcoholic Fatty Liver Disease Using Intelligent Methods

Abstract

Nonalcoholic Fatty Liver Disease (NAFLD) is the most common cause of chronic liver disease around the world. Remaining silent in the early stages makes its evaluation a challenge. Liver biopsy is still the gold standard method used to classify NAFLD stages but has important sample error issues and subjectivity in the interpretation. This research is an effort to overcome liver biopsy to a possible extent by forming a non-invasive clinical spectrum. This paper proposed an intelligent scheme using the forward algorithm, Viterbi algorithm, and Baum-welch algorithm for examining the disease, and a new clinical spectrum is introduced that incorporates most likely attributes associated with NAFLD stages. The experimental results verify that our method is efficient in distinguishing the credibility of an attribute being associated with a specific stage in case it is linked with more than one stage. Moreover, the proposed scheme can successfully estimate the likelihood of stage progression and supports medical knowledge more proficiently and realistically.

Keywords

• Nonalcoholic fatty liver disease
• computational methods
• forward-backward learning
• intelligent systems
• healthcare informatics