<![CDATA[Laboratory tests of blood and urine chemistry are important in determining a patient's health status. The results of these tests are usually presented in the form of a table that lists medical abbreviations alongside corresponding values. Medical professionals can easily interpret the results by determining whether the values are within the normal range or indicate an abnormality. However, junior doctors may still require reference tables to compare each component's value obtained from the laboratory test with the normal range. Therefore, this study proposes a Natural Language Generation (NLG) approach using the Roulette Wheel and Bigram algorithms to assist junior doctors in efficiently and effectively interpreting blood and urine chemistry test results. This system will convert the numerical data into Indonesian text, which will become the narrative interpretation in the laboratory report. Evaluation by junior doctors and medical professionals showed a naturalness level of interpretation of blood and urine chemistry test results ranging from 86% to 96%]]>
Copyrights © 2023
