Related Experiment Video
Updated: Aug 14, 2025
Automation of the Micronucleus Assay Using Imaging Flow Cytometry and Artificial Intelligence
Published on: January 27, 2023
The development and validation of a prediction model of lithium carbonate blood concentration by artificial neural
Jihua Zhang1, Yanping Wang1, Wenying Zhang1
1Institute of Mental Health, Tianjin Anding Hospital, Mental Health Center of Tianjin Medical University, Tianjin, China.
Background:
Bipolar disorder (BD) is common in clinical practice. Lithium (Li) carbonate is often used in the treatment of BD. However, the therapeutic dose of Li carbonate is close to the toxic dose, and Li poisoning is prone to occur. Precise prediction of Li concentration will help clinician to identify patients at high risk of toxic dose of Li carbonate. The purpose of this study was to establish a model for predicting the blood concentration of Li carbonate through an artificial neural network (ANN), and to provide a basis for the clinical rapid and effective formulation of individualized dosing regimens.
Methods:
Patients with BD who were diagnosed and treated in our hospital from October 2016 to April 2021 were enrolled as the research participants. We collected patient demographic data, including age and gender; physical examination information, including height and weight; laboratory test results, including liver and renal function, and Li concentrations; medication information, including Li carbonate usage, concomitant medications, and dose; and information on comorbidities and adverse reactions. The Li concentration data of 236 patients were randomly divided into 2 groups: 195 cases in the training group and 41 cases in the test group. The ANN fitting module of SPSS 26.0 was used for modeling and prediction.
Results:
A total of 236 patients with BD were included in this study. Daily dose (before testing Li concentration), age, alanine aminotransferase (ALT), aspartate aminotransferase (AST), albumin (ALB), total bilirubin (TBIL), and creatinine (Cr), and co-administered zopiclone, quinine, tipine, lorazepam, olanzapine, valproate, metoprolol, and statins were used as model input variables for training. The test results of the model in the testing group showed that the correlation coefficient between the predicted value of Li concentration and the actual value was r=0.9883, r2=0.9767, P<0.001, the prediction error range was -0.05 to 0.07 mmol/L, and the deviation range was -18.52 to 13.04%; the mean absolute error was 0.03, and the mean prediction error deviation was between -10% and 10% in 33 cases (80.5%).
Conclusions:
The correlation, accuracy, and precision of ANN prediction are worthy to be further investigated to predict the blood concentration of Li carbonate.
